^ General Scheme of Electoral Reform Bill 2020: Discussion (Resumed)

It is great to be able to see people face to face in the committee. We have been used to seeing each other on screens for a long time, so it is great to have the opportunity to meet face to face in public.

Today, we are continuing our pre-legislative scrutiny of the general scheme of the Electoral Reform Bill 2020. After this meeting we will suspend for 15 minutes for our private meeting on correspondence. Apologies for his absence have been received from Deputy Duffy.

We are joined remotely by the witnesses, Mr. Dualta Ó Broin, head of public policy in Facebook Ireland Limited, and Mr. Ronan Costello, senior manager of public policy in Twitter. Their opening statements have been circulated to members. I will invite our witnesses to make their opening statements and then members will be invited to ask questions. Members' questions and answers will be limited to five minutes.

On privilege, members attending from their Oireachtas offices are protected by absolute privilege in respect of their participation in this meeting. This means they have an absolute defence against any defamation action for anything they say at the meeting. For witnesses attending remotely, there are some limitations to parliamentary privilege and they may not benefit from the same level of immunity from legal proceedings as a person who is physically present. I remind members of the constitutional requirement that members must be physically present within the confines or place at which Parliament has chosen to sit, Leinster House or the Convention Centre Dublin, in order to participate in public meetings. Witnesses and members are directed that only evidence connected with the subject matter of these proceedings is to be given and are asked to respect the parliamentary practice to the effect that, where possible, they should not criticise or make charges against a person or entity by name or in such a way as to make him, her or it identifiable. The opening statements submitted to the committee will be published on the committee's website after this meeting.

I invite Mr. Ó Broin to make his opening statement on behalf of Facebook Ireland Limited.

Mr. Dualta Ó Broin

I thank the committee for inviting Facebook to contribute to the pre-legislative scrutiny of the general scheme of the Electoral Reform Bill 2020. I am head of public policy for Facebook Ireland Limited. I will start by saying that Facebook very much welcomes the Government’s intention to establish an electoral commission and its objectives of modernising electoral law and safeguarding election integrity. As members may be aware, Facebook has made significant changes to its services to increase transparency, to prevent interference and to remove misinformation in the context of elections. In addition, Facebook continues to collaborate with organisations worldwide that are working to strengthen democracy, good governance and election integrity.

We recognise that the desire to introduce greater regulation and transparency in online political advertising has been a focus in Ireland for several years. The introduction of the Private Members’ Bill on the subject in 2017 led to the Open Policy Forum on Political Advertising in 2018, in which we were very pleased to be invited to participate, in addition to the consultation at the time. It is also important to highlight the role which Irish parliamentarians have played on the International Grand Committee on Fake News and Disinformation, working for the past few years with colleagues in parliaments across the world to identify solutions. From an Irish perspective we were particularly pleased that our vice president of content, Monika Bickert, was invited to attend the session held in the Seanad in November 2019.

Facebook has reviewed and analysed the general scheme with great interest but also, at times, with some concerns. Our detailed written submission was provided to the committee on 23 February. In that submission, we summarised Facebook’s efforts to protect democracy and election integrity, before noting our views on the draft provisions concerning online political advertising. In this statement, as requested, I will focus only on the regulation of online political advertising, which is dealt with in Part 4 of the Bill. For ease of reference, we have included a summary table of our position on the heads in Part 4 in annex 1 to this document. In addition, as it is likely to be of interest to the members, we have included the executive summary of our submission to the media committee on the Online Safety and Media Regulation Bill 2020 as annex 2. Given the amount of engagement that has taken place on the issue in Ireland, we would be interested to see the regulatory impact assessment which was prepared on this aspect of the Bill and, in particular, the reasons the approach outlined in Part 4 of the general scheme was deemed optimal.

As an overarching point, it is Facebook’s belief that online political advertising rules and standards should be harmonised across the EU to the extent possible. We are concerned that enacting these provisions at this time will lead to a lack of alignment between the requirements which exist during electoral periods in Ireland and the year-round EU wide rules for political advertising which the European Commission intends to introduce via the European democracy action plan and the Digital Services Act. We understand that the Commission intends to publish a legislative proposal for political advertising in the next six months. There is, therefore, a real possibility that online platforms could find themselves under two inconsistent regulatory regimes and placed in the undesirable, and avoidable, situation where they would have to choose between violating Irish law or violating EU law.

Second, if the committee is of the view that this part of the general scheme should rapidly progress, despite the concerns expressed by Facebook and others, we strongly recommend that a principles-based approach is adopted. This would give the electoral commission the power to engage with companies and other stakeholders to promote ongoing innovation and improvements, while at the same time delivering on the Government’s stated objective of achieving transparency in online political advertising.

Regarding the specific wording of the proposed legislation, it seems appropriate to start with the definition of a political advertisement. Facebook recommends a more objective definition of online political advertising. At a minimum, the legislation should take account of what online intermediaries such as Facebook can and cannot do. In short, we can identify whether an advertisement contains certain content, but we cannot divine the intention that an individual has in placing an advertisement. The reference to "political purpose" should be replaced with a far more objective test.

Moving to the actions required of online platforms, we have a broad range of concerns with the requirements set out in respect of the transparency notice. These include privacy concerns about the scope of information regarding an individual which is intended to be shared and the impact of disclosing specific micro-targeting criteria. We also have practical concerns about the appropriateness of online platforms collecting information about the amount spent on content creation and the requirements for advertisers to provide an end date for a campaign. Finally, we have concerns about the reporting obligations set out in this head and question whether they are proportionate or indeed necessary in all circumstances.

The appointment of a responsible person deserves significant attention. The general scheme provides for a complex system of oversight to be overseen in the first instance by a responsible person appointed by the online platform. The appointment of the responsible person moves the responsibility and onus of providing accurate and complete information away from the buyer, who holds the information, and onto the online platform.

Our concerns in this regard are far-reaching. It seems far more straightforward for the relevant regulatory body to have the power to simply engage with the online platforms and directly investigate any and all of the issues which these heads appear to be attempting to resolve.

As a final point, I know that a broad range of issues which are not currently captured by the general scheme, such as misinformation and cybersecurity, have been discussed during the deliberations of the committee. In light of the volume of regulation for online services which is emerging at EU level and the existing and new Irish regulatory bodies which may be required to implement these measures, it may be of benefit to consider how the proposed electoral commission would interact and engage with some of these, including the digital services coordinator which will be required under the digital services Act, as well as Ireland’s existing cybersecurity agency.

I again thank the Chair and members for inviting me to speak. I look forward to the discussion and hope to be able to answer as many questions as possible from members.

Mr. Ronan Costello

I thank the committee for its invitation to participate in this session. I am a senior public policy manager for Twitter in Europe. I will first provide our observations on the general scheme of the Bill and then outline some key elements of our election integrity work.

Twitter is an open and public platform. Our singular mission is to serve the public conversation. The public conversation on Twitter is never more important than during elections. We are a global service and we strive to think globally, so every year is an election year. We take what we learn from each election cycle and use it to improve our election integrity work worldwide. Policies that were developed prior to the European Parliament elections in 2019 were improved for the US Presidential election in 2020, and on it will go through 2021 and beyond. We daily build on our efforts to protect the public conversation and enforce our policies against deliberate attempts to mislead people and manipulate our platform. Partnerships with industry peers and civil society, as well as political parties and government officials, are critical to our success and it is in this context that we welcome today’s dialogue.

Twitter shares the fundamental objective of the Bill. We want to make elections more transparent, encourage accountability and promote an honest and informative civic conversation. We welcome the proposed establishment of an electoral commission. In countries where we have worked with such a commission, such as the UK, our experiences have been very positive. Having a single point of contact is helpful for everyone involved. During the European Parliament elections, we channelled our engagement through a dedicated election team. When the process concluded, the director-general of the directorate for communications in the European Parliament reached out to express appreciation of this work.

As regards the definition of "political purpose" currently provided in the Bill, there is a remote chance that the day-to-day campaigning of non-profit organisations and NGOs may be captured in the scope of the legislation. Additionally, this definition may capture apolitical voter education and engagement campaigns that the Department of Housing, Local Government and Heritage or the newly-established electoral commission would seek to run. We respectfully submit that such categories of promotion should not be subject to the requirements set out in the draft legislation. This would likely bring unforeseen and unintended consequences.

In October 2019, Twitter decided to prohibit political advertising outright. We took this step based on a belief that political reach should be earned, not bought. We cautioned that the effects of algorithm-optimised and micro-targeted political messaging present new challenges to civic discourse which are not yet fully understood.

As regards Twitter and elections generally, prior to an election the public policy team co-ordinates a period of preparation. We connect with political parties to provide an overview of our policies. We work around civic integrity, as well as sharing practical tips with them for staying safe on Twitter. We also engage with parties and the relevant electoral authorities to ensure the right support channels are in place to escalate concerns and report any perceived violations of our rules. This external outreach is complemented by the gathering of a cross-functional internal team of experts who work on elections.

Our civic integrity policy is a recognition of the responsibility to protect election conversations from interference and manipulation. Under the policy, we prohibit attempts to use our service to manipulate or disrupt civic processes, including through the distribution of false or misleading information about the procedures around voting. Core to our work around civic integrity are our efforts to strengthen Twitter against attempted manipulation, including malicious automated accounts and spam. The fight against malicious activity and abuse has been significantly augmented by our continuously improved capacity to leverage machine learning to recognise these harmful practices. Through machine learning, we can detect behaviour-based signals that help to scale our moderation efforts. Our capacity to deploy these techniques now means that more than one of every two Tweets that is removed for violating our rules has been proactively identified without requiring user reports.

We also recognise that addressing harms associated with misinformation requires innovative solutions. Content moderation in isolation is not particularly scalable. This is why we have sought to be creative in pioneering new approaches to labelling harmful content and experimenting with community moderation. Labels help us to clarify that a claim may be disputed and then we can link to more authoritative sources.

Finally, Twitter supports regional and global regulatory alignment around technology legislation. We urge consideration of existing initiatives that have provided a structure for regional standard-setting and co-operation, including the EU code of conduct on disinformation, the upcoming democracy action plan and the digital services Act. As debate around the world focuses on how to solve public policy challenges related to our online environment, the approach of Twitter to regulation and public policy issues is centred on protecting the open Internet and we welcome any dialogue with policymakers that advances this goal.

I thank Mr. Costello. I will now move to members. I call Deputy McAuliffe.

For full transparency, I wish to state that I use both of the products being discussed and have paid for political advertising, as, I am sure, have many politicians. The Bill seeks to regulate political activity during election periods. My first question is for Mr. Ó Broin in particular as Facebook permits political advertising. Does it believe that a year-round system would be more appropriate? I ask Mr. Ó Broin to address the creation of custom audiences and the SIPO requirement to apportion costs spent during the year in terms of being reused during a campaign.

Mr. Dualta Ó Broin

I thank the Deputy. The rules we have in place apply all year round. We try to provide as much transparency as possible. A related point regards a far broader category of issues than what seem to be captured by the definition in the Bill. On the point around SIPO, I am not entirely sure what the question is, but we provide-----

I will rephrase the question. Through micro-targeting and spending on advertising pre-election, it is possible to create custom audiences within Facebook which then can be used during the election period. As such, in reality, unlike other media such as print advertising or radio advertising which have their benefits immediately during the campaign, online advertising offers a different opportunity to create audiences that then can be used during a campaign.

Mr. Dualta Ó Broin

The way our systems work is that one is able to target advertising based on certain categories. It is the case that one is able to create a category of interest that one would use to target specific advertising. However, coming back to the point around spend, etc., what we show on the ads library is the amount that is spent on the ad in delivery, that is, when it is being seen by people on the platform. We do have a point around the whole issue of content creation which seems very difficult to capture for us as a platform. That may be something that individuals or political parties could report directly to SIPO.

I take it from the submission provided by Mr. Ó Broin's that the preference of Facebook would be for year-round regulation. My argument is that it would strengthen the Bill if the regulation were year-round because of the nature of it.

Mr. Ó Broin raised concerns regarding privacy, the collection of data and the engagement of the commission with Facebook rather than it being transparent. I would argue that a voter has not just the right, but the responsibility to see what is happening during an election, and that it would not be enough for the commission to be able to engage with Facebook. It needs to be something that is visible. If I see a leaflet with which I disagree, I should have the right to see the source of the leaflet. Equally, the same should apply for Facebook.

Does Mr. Dualta Ó Broin agree that the personal details of anyone participating in an election campaign should be displayed?

Mr. Dualta Ó Broin

Coming back to the overall point, we agree that any political activity or activity that is seeking to promote views or pay for content relating to a broad range of social issues should be seen. That is why we have built the ads library that applies year-round. In respect of the detail about an individual, it raises privacy concerns. I know that the deputy commissioner appeared before the committee last week and he said it would be possible to build a system that was safeguarded and put processes in place to mitigate the risk. That work on the data impact assessments required may be ongoing in the Department. Based on what we can see in the general scheme, it raises privacy concerns. We believe there is a more straightforward way of going about this which would not necessarily have the same level of privacy concerns. One, for example, would be that an individual-----

If someone competes in the public space in a political campaign, surely people have the right to know who is behind those campaigns.

Mr. Dualta Ó Broin

Fundamentally, it will be for the Deputy and his colleagues, as legislators, to decide where the balance is and to ensure that the right privacy protections are in place. We are saying there is a more straightforward way of doing it. We will provide the name of the person and all the details regarding the ad, but it may be more appropriate to provide an individual's home address directly to the regulator. We have seen examples in recent referendums where there have been highly emotionally charged debates where it might not always be in the public interest to have an individual's home address available online. It is just a question of balance and we are just saying that the home address issue raises privacy concerns. It obviously it will be a matter for the legislators to decide where the balance is in that instance.

Does Mr. Dualta Ó Broin accept that we seem to have different rules for different media? This Bill permits providing online political advertising to a content provider, which clearly states it is not in the news space, yet we ban political advertising on radio stations, which are often very good-quality local and authoritative sources of news. I believe there is a discrepancy there. Following what Twitter has told the committee, does Mr. Dualta Ó Broin believe there is an argument to ban online political advertising?

Mr. Dualta Ó Broin

First, we recognise the value of authoritative sources of news and that is why we work with trusted flaggers across the globe to combat misinformation. We do not believe that banning political ads is the answer. We believe that greater transparency in political advertising is the answer. Can an argument be made regarding radio and television? They are different types of media and perhaps the Broadcasting Authority of Ireland would have a view on that.

My questions are for Mr. Dualta Ó Broin. Facebook's submission to the committee is very disappointing. The proposition in the legislation is incredibly modest. All it is proposing is that for the three or four weeks of a general election campaign, election ads or ads that are designed to have an impact on the electoral outcome should be designated as such; information about the way in which those ads are being targeted and funded should be made public; and there should be an onus on the platform that publishes these ads to take responsibility and ensure that the advertiser is compliant with the law. Those are basic propositions. I am concerned that Facebook is proposing to the committee watering down what is already a very weak proposal. I want to put that on the record.

I agree that we need to get the definition of political advertising right, as NGOs and the Irish Council for Civil Liberties have highlighted. I do not see how Mr. Dualta Ó Broin's definition addresses any of those issues. In fact, subsection 3 includes most of the areas about which NGOs and human rights organisations would seek to be able to advertise. Why does he believe his definition, as outlined in annex A, is better than the Government's? In other words, what could the Government's definition capture that his does not?

At no point in the general scheme are data protection concerns mentioned, nor does it mention anybody's home address. It refers to a name and postal address, but election agents and election posters often use an office address. Some of us might have used our county council addresses when we were elected councillors. In the same way that candidates' and-or election agents' information is made public, I do not see why it should not be the case here.

Mr. Dualta Ó Broin made a spurious argument by stating that the regulations on postering are different and less onerous than microtargeting. However, when a candidate spends five times more money on posters than anybody else, we see the posters on the streets and the lampposts whereas there is not public visibility on microtargeting. Therefore, I do not see how that is a valid concern. I ask him to give a bit more detail on that.

My biggest concern relates to Mr. Dualta Ó Broin's comments about head 122. He seems to suggest that rather than Facebook having a responsibility to ensure that people, who place ads on its platform and from whom it profits, comply with the law, it should be the responsibility of the electoral commission. Only when there is - I quote the Facebook submission - "persistent refusal or non-compliance with warnings" would it reach an escalation. That suggests that people would be able to break the law and persistently be non-compliant. Facebook would profit from it but at some point, because of its warnings, it would escalate to the electoral commission. That seems to absolve Facebook of any responsibility for the content on its platform. I do not see how having a person responsible is anything other than a natural safeguard.

Mr. Dualta Ó Broin keeps describing Facebook as an intermediary, but ultimately his company is profiting from the ads that any of us places. Like Deputy McAuliffe, I place ads on Facebook. As Facebook is profiting from the ads we place, there should be an obligation. I ask Mr. Dualta Ó Broin to address some of those concerns and I may come back in the next round if there is time.

Mr. Dualta Ó Broin

I do not believe that what we are arguing for is a watering down of the provisions in the legislation. I think we are arguing that the provisions should be more effective and should give the regulator the appropriate tools with which to engage with companies such as ours to establish the practices that are in place and also to be able to react and adapt to any new threats coming down the road.

Regarding our approach to the definition of social issues and electoral political ads, I had a quick look at the ads library this morning. It contains ads from NGOs and from companies dealing with sustainability, sustainable energy etc. We believe all that influences the democratic process here and that is why we require those individuals and organisations to go through those steps and measures to put it into the ads library so that everyone can see what the ad is, who is paying for it, the categories of targeting that were used, what the spend range is and other details on the ad.

That is the point on definition. How that is different from what the Government is suggesting? It is not clear to us what a political purpose ad is. It is one thing to say it contains content about social issues, electoral issues or political issues and that is a far more objective test than trying to figure out what the intention of somebody behind the advertisement is. That is the point we are making. It is moving from the subjective to a more objective test.

On the point of microtargeting, the releasing of the targeting criteria is the main thing academics have been calling for since 2016 or 2017. We have real privacy concerns over releasing the targeting criteria within the EU. However, we are not saying that we have privacy concerns, and that we are just throwing our hands up and saying we can do nothing about it.

We have been engaging with the European Commission in recent years and with academics at a European level to figure out what would work-----

Will Mr. Ó Broin be more explicit about what those privacy concerns are? It is not clear from his submission.

Mr. Dualta Ó Broin

If the exact targeting data for an ad are released, it might be possible for bad actors to work back from the targeting criteria and the audience to figure out who the individuals were. At a high level, that is what the concern is. We are working with the European Digital Media Observatory, which was set up by the European Commission as part of its efforts regarding the code of conduct on disinformation, to figure out what a code of conduct would look like under the general data protection regulation, GDPR. To return to what the deputy data protection commissioner, Dale Sunderland, asked last week, what are the frameworks, protections, risk mitigation measures and processes that need to be in place to release this information? That work is ongoing. We are not averse to doing it. We have released the data in the US and I can share the link for that, but we have a concern about privacy.

I thank both witnesses for their attendance. I have read with interest, as I am sure all members have, the points made by the witnesses in their submissions and there are some good ones in respect of harmonising EU and Irish law in this area. While that should certainly be considered, it is not a reason not to have more transparency in regard to online advertising and to where the money for that advertising comes from. In an earlier response, Mr. Ó Broin stated that greater transparency is the answer. I compliment Facebook on its ad library because it introduces an element of transparency to political advertising in this country by identifying where individuals or organisations that run adverts on behalf of parties or politicians are located. Although we know the location of these administrators, however, we have no idea who they are, how much they spend on behalf of parties or politicians or where the money comes from. As we heard at last week's committee meeting, the Facebook ad library was designed for the US, where it is okay for individuals or organisations to spend unlimited sums on behalf of politicians or parties, but that is not okay in our jurisdiction. I do not believe that spend is being captured on the ad library.

I have a few short questions on which I would appreciate yes-no answers. Will the representative from Facebook confirm that if I approve an individual or organisation to place adverts on Facebook on my behalf, this ad spend is not accounted for on my ad library but only on the ad library of the person I have approved?

Mr. Dualta Ó Broin

The page will be authorised-----

Mr. Dualta Ó Broin

It links back to the page-----

Will Mr. Ó Broin confirm that if I approve an individual or organisation to place adverts on Facebook on my behalf, that ad spend is not accounted for on my ad library page?

Mr. Dualta Ó Broin


Mr. Ó Broin is saying it is not accounted for on my ad library page but on that of some unknown individual. We know what country they are in but we have no idea who they are. Facebook knows who they are but the public does not. Is that correct?

Mr. Dualta Ó Broin

No. To be clear, one of the administrators on the page must be authorised to run political ads in Ireland, and in order to do that, they will have their identity and address verified and we will ensure they are in Ireland. The Senator is talking about another page-----

I am asking where the ad spend is accounted for. Is it accounted for on the page of the administrator I have approved or on my ad library page?

Mr. Dualta Ó Broin

It is on the second page to which the Senator referred.

It is accounted for on the second page and, as a member of the public, I have no idea who that person is. I know there is an account but it could be in Serbia or Germany and I have no idea who is behind the page.

I have a couple of follow-on questions-----

Mr. Dualta Ó Broin

I wish to clarify that point because it is very important. That other page is also required to go through that verification process. It will have an administrator who is verified to run ads in Ireland with their Irish identity and Irish address. An administrator needs to be in Ireland to run political ads in Ireland.

I am glad Mr. Ó Broin moved on to that. He indicated that if I have approved individuals or organisations based in other countries to manage my page, they cannot run the advert, which I accept. Could they use virtual private networks, VPNs, to circumvent that? Could they create the campaign, develop the ad sets, create the targeting, examine the demographics and create the content while somebody in Ireland who is also linked to my page would press the run-advert button on Facebook?

Mr. Dualta Ó Broin

From speaking about what-----

I just want to know whether it is possible.

Mr. Dualta Ó Broin

I can speak about what happens on our platform. I cannot speak about activity that takes place before somebody starts to create an ad on our platform or decides to publish an ad on our platform.

That is not what I am asking.

Mr. Dualta Ó Broin

From creation through to publication of the ad, only the person who is authorised and in Ireland can be involved in that on our platform.

A person who is based in another country cannot create an ad set or campaign or examine the demographics for targeting on Facebook. Why would somebody in another country be linked to my ad account if they can do nothing on the platform?

Mr. Dualta Ó Broin

There is a broad range of activities in which page administrators are involved that go beyond political advertising. We have a specific set of requirements for political advertising. Page administrators can be involved in moderating discussions on a page using the safety and security tools we provide or creating organic, or unpaid, content. They are involved in a range of activities outside of political advertising.

To be clear, is Mr. Ó Broin saying that somebody who is based in another country cannot use Facebook's ad creation tools to create a campaign or ad sets, examine demographics or create content?

Mr. Dualta Ó Broin

It would be a violation of our policies-----

Is it possible? I return to the original question. It is very important in this context.

Mr. Dualta Ó Broin

On our platforms, to be involved in the creation of an ad or in deciding where the ad will be, what its targeting criteria are or to publish the ad, that all has to be done by an individual who is-----

Pressing "publish" on the ad has to be done in Ireland but is it possible for any of the other steps to be taken in another country? I apologise for pressing the issue but it is very important in the context of the transparency aspect of this legislation. That is what we are trying to capture in heads 123 to 125, inclusive.

Mr. Dualta Ó Broin

Obviously, I cannot say what happens off the platform. That is not something I can comment on. I can comment on what happens on the platform.

I am not happy with the response because it is not clear.

If we stick to the time limits, we will have time to return to the issue. It is an important aspect and we should ask about it.

I would appreciate that.

I thank Twitter and Facebook for allowing their representatives to appear before the committee to engage and co-operate with us because that is really important. This is perhaps one of our most important meetings because we are talking about electoral reform and the importance of election integrity, confidence and trust.

This is the cornerstone of democracy, locally, regionally, nationally and internationally. We know that these platforms are in demand. I will not share too much detail with the committee, but I took the time to check out the Twitter and Facebook accounts of all Oireachtas Members. What I found is very interesting. Many members will be familiar with the type of engagement that happens and many people will have been impacted upon negatively by that engagement. It is significant that representatives of Facebook and Twitter are with us this morning.

The public conversation regarding Twitter and Facebook, particularly in the context of elections, is fundamental. Content influences decisions, be they good, bad or indifferent, based on misinformation, false news or wrong motives. That is the reality. While Facebook and Twitter are unique and powerful, they are also potentially very powerful instruments to be used to abuse people and the political process. That has to be a concern for anyone who represents the democratic parliament. Every day, we are having these conversations, but particularly in an election, they determine the outcomes. That is a really important point.

I would like to ask two questions because I want to hear more from Facebook and Twitter. They arise on foot of the opening statements. My first question is to Mr. Ó Broin. He mentioned the danger of online platforms such as Facebook finding themselves under two inconsistent regulatory regimes in terms of EU and Irish law and stated that Facebook has been dealing with this issue in the context of Brexit. I am familiar with Brexit and the challenges it poses for the North, the South and the Border area. Surely, Facebook has found itself in a similar situation in the past. If so, how did it handle it?

My next question is to Mr. Costello. He rightly stated that we have not fully realised the dangers of the algorithm-optimised and the micro-targeted political messaging on social media platforms, as referenced by previous speakers. In Mr. Costello's opinion, what steps could a body such as the proposed electoral commission take to offset some of these negative affects and the phenomenon in an Irish context? I would appreciate it if the witnesses could give me relatively concise responses.

Mr. Dualta Ó Broin

The point we are making is that the EU has said it proposes to legislate in this area and Ireland is taking quite a prescriptive approach to transparency of political advertising requirements. Given that it is only six months away, it seems more prudent to us to hold off to see what the requirements at EU level will be. At that point, the electoral commission or the Houses of the Oireachtas could decide whether they go far enough. Electoral law will always be a member state competence. It is that question of alignment, particularly in view of the fact that the European Commission has said it proposes to provide for transparency in political advertising across the EU.

I thank Mr. Ó Broin and I invite Mr. Costello to respond.

Mr. Ronan Costello

On the electoral commission and how we will engage with it, I mentioned earlier that Twitter has a civic integrity policy. Part of it is to counter the proliferation of disinformation with respect to the electoral process. If the electoral commission is the entity for Ireland that verifies results, counts and so on, Twitter would, for example, take its lead from the commission if someone were to tweet that a count had gone one way when it had actually gone the other or that a result had been declared in favour of one candidate, whereas the electoral commission had declared in favour of another. Based on the electoral commission's determination, we could then take it as an authoritative source in considering whether to label tweets and, through that labelling, whether to direct our users to more authoritative sources such as the electoral commission or a credible news platform. That is one way we could foresee Twitter working with the electoral commission. Another parallel might be the work we are doing with health organisations around Covid-19, where we have escalation routes for them to send us links to content that they believe is violating Twitter policy on harmful misinformation with respect to Covid-19. An electoral commission could have a similar relationship with Twitter, where it would have an escalation route to us and could share with us what it believes to be violations of Twitter civic integrity policy around information that is misleading people in the context of how they engage with the electoral process and the outcome of electoral votes.

I thank Mr. Costello. I have a couple of questions arising out of submission from Facebook, in particular annex 1. Mr. Ó Broin said Facebook has difficulty with the definition of "political advertisement" and it is seeking a stricter definition to be applied to it. I note from the Twitter submission that it has a comprehensive definition of "political content". Would a tightening up of the definition of "political advertisement" satisfy the concerns of Facebook in that regard?

Mr. Dualta Ó Broin

If the test is more objective, it would. That is the key issue for us.

Twitter appears to be able to be objective in defining what constitutes "political content". Facebook could apply that same objectivity based on the definition applied by Twitter and inclusive of the suggested definitions of what it should include, as set out by Mr. Ó Broin. It is an issue that can be overcome in the context of this Bill.

Mr. Dualta Ó Broin

Absolutely. As an overarching point, we believe the Bill should progress and the electoral commission should be established. I do not think there is anything insurmountable in the concerns we have in regard to online political advertising. We can come to a solution on it.

I thank Mr. Ó Broin. We will now move to head 122 and the concerns regarding the appointment of a responsible person. The term "responsible person" as set out in the in the heads of the Bill is a term already provided for in the Electoral Act as being a person in authority within either a third party or the accounting unit of a political party. Its inclusion in this Bill may be considered necessary to ensure consistency with the existing electoral code. As I said, the term "responsible person" already exists in the Electoral Act. The concern of Facebook appears to be that there are substantial privacy and commercial confidentiality issues with the appointment of that responsible person if it is a third party. It could be included in the contract advertisement that the responsible person will be appointed and it may be a third party? It is possible to overcome this issue as well.

Mr. Dualta Ó Broin

We would have more serious concerns regarding the responsible person than we would have in the context of other aspects of the general scheme. Where an advertiser provides information to us, we can provide that to the regulator, and if there is any question or cause for concern on the part of the regulator it can follow-up and investigate with the advertiser.

Sure, but we are talking about a short timeframe in this legislation in terms of the couple of weeks of the electoral cycle. If we were to go with Mr. Ó Broin's suggestion that a responsible person should be just a point of contact, namely, a mere conduit that bears no responsibility. There has to be a responsible person appointed. Mr.Ó Broin spoke about developing a reporting mechanism where persistent refusal or non-compliance would trigger greater enforcement. He must be aware of the timeframes involved here. Surely, we would be outside of the electoral cycle if we were to go with what is suggested by Facebook.

Mr. Dualta Ó Broin

We have concerns about the short timeframe within which these provisions would apply. We have touched on that already. It comes back to the question of providing the transparency and putting the material into the ads library such that not only the regulator but all stakeholders can engage with it and see what is taking place. We are happy to report to the regulator as required. When it comes to any question on inconsistencies or, as mentioned by others, sources of funds and so on, we see that as a role for the regulator to play, not the platforms.

Mr. Ó Broin has an issue with the establishing of a business relationship as outlined in head 123. He is concerned about what happens if he already has a business relationship. If there is an existing business relationship, which then tries to submit political advertising, that would trigger this aspect of the Bill. There can be an existing relationship but if it gets into political advertising, as defined, then it would trigger that aspect of the Bill. Does Mr. Ó Broin think that is a way of dealing with this issue?

Mr. Dualta Ó Broin

To come back to where we are coming from on these provisions, if the principles are set down in the legislation for what the regulator is tasked with doing, the regulator is then empowered to engage with platforms such as ourselves to ensure we are complying with those principles. For example, the legislation provides for transparency in online political advertising. Many of the issues that arise in heads 121 to 125, inclusive, are then avoided, but two things are still ensured. The objectives of the State are met because one is ensuring there is regulatory oversight of the transparency of online political advertising but also that the regulator can evolve its approach, and can require companies to evolve their approach, in the face of emerging threats.

I thank Mr. Costello and Mr. Ó Broin for coming in, for participating and for their opening statements. Like other committee members, I read the book by Christopher Wylie, who was a whistleblower for Cambridge Analytica, Mindf*ck, a couple of years ago. He clearly detailed his experience working for Cambridge Analytica in respect of misinformation, voter turnout suppression, campaigns and the targeting of sections of society already under-represented because of voter suppression campaigns, which is a terrible use of technology and social media to target people who are already disempowered. The end result of that misuse of technology and social media platforms is that people were less empowered again following the electoral process. Mr. Wylie was doing this work in 2013 and 2014, so it is not particularly new. He went through the detail of how micro-targeting was used to build audiences and heighten their fears, anxieties and prejudices over time. I recommend that anyone who has not read the book read through it or even read The Guardian articles he wrote about it.

My question is specifically for Facebook. Given there have been terrible consequences from the misuse and abuse of micro-targeting, which has been done through social media platforms, is the cleanest, safest way to deal with safeguarding electoral processes, and safeguarding against manipulation and voter suppression campaigns, to simply ban micro-targeting during political debates, campaigns and elections? What is Mr. Ó Broin's view on that?

Mr. Dualta Ó Broin

As a company, we have learned an enormous amount since the timeframe referenced by the Deputy. We do not agree that banning micro-targeting is the correct approach. We agree that ensuring there is transparency in what is happening in spending on political advertising and, in our case, social, electoral and political issues, is the correct approach to take. As I mentioned in a previous response, we are committed to finding a way to work with academics to release more of that data once a framework has been established to do so within the EU.

Much of the discussion on targeting is on the negative side. I ask that consideration be given to the upsides of targeting in political campaigning. We have seen incredibly positive uses of our tools and services across the world through ensuring that all players have access to tools to spread their messages to interested audiences.

All of us will probably agree that the role of social media in all the information, debate and discussion during election and political campaigns is very important. Some of the very sinister practices that have occurred over previous years have been around micro-targeting. Mr. Ó Broin referred to the positives of micro-targeting in identifying specific audiences and giving them specific information and messages that other audiences cannot see. He might elaborate on why he feels there are positives in micro-targeting and why transparency alone is enough to deal with that. Can he give some positive examples of micro-targeting because I have outlined some of the negatives around it?

Mr. Dualta Ó Broin

Certainly. At a geographic level in Ireland, we have constituency boundaries. It is obviously of benefit for candidates to be able put a message into a particular constituency. That is just one example.

There is a difference between targeted advertising and micro-targeting. If I take out a targeted advertisement in my local newspaper, it can be seen by anyone, from a small child starting to read up to the oldest voter. It is targeted but it is not micro-targeted. I am not picking people out based specifically on age, demographics, class, different interests and so on. There is a difference. I have no issue with targeting. Of course, it makes sense for people to get an advertisement in the Dublin region, or in a Dublin or Waterford constituency, but there are concerns about the picking out of particular segments, and the manipulation and lack of transparency around that.

Mr. Dualta Ó Broin

We do not see a lack of transparency because all the advertisements are in the ads library. If one takes the health messaging at the moment, different messages are appropriate to the over-75s compared to the under-25s when it comes to Government messaging on the health measures currently in place. There are positive, totally appropriate ways of using targeting based on demographics.

Most issues have been covered by previous questions, but I will make an observation on the Bill. From the perspective of the social media companies, Liz Carolan, who was in front of the committee last week, and academics such as Dr. Jane Suiter, the one issue that comes across is that the Bill is, perhaps, too limited and too prescriptive. We need a dynamic electoral commission, involved and able to keep up with technology and the different methods more nefarious actors involved in online advertising will be able to exploit going forward. In that respect, the heads of Bill are perhaps too limited. As Deputy Ó Broin stated, it is quite a conservative change. I am not sure it will be able to keep up with the changing face of online political advertising.

I agree with Facebook when it states the Bill is not fit for purpose, but perhaps not for the same reasons.

Within the longer submission supplied, there was talk of privacy concerns about identifying natural persons. Last week, we questioned the deputy Data Protection Commissioner who said he considers that the definitions involved in that do not give rise to any privacy concerns. Perhaps Mr. Ó Broin can outline why he thinks that is the case. I am not sure if he saw the representatives from the Data Protection Commission.

In terms of aligning this legislation with the European Commission's regulations that are coming down the road, what parts do the witnesses think we should include in the Bill? Can they suggest a type of electoral commission that is outside the terms of this Bill? What principles would they prescribe in the Bill for an electoral commission? I ask because people across the board have identified the necessity for the electoral commission to evolve as opposed to how it is in this legislation.

Mr. Dualta Ó Broin

On the privacy point, yes, I had the opportunity to see the session last week, and we both have had the opportunity to see previous sessions as well. As a general point, I thank the members for the amount of time that they are giving to the pre-legislative scrutiny of this legislation. To come back to privacy concerns, we are looking at a general scheme which we believe requires identification. We said there are privacy concerns in terms of that, and we outline that it is not, in our view, appropriate to legislate in this space until all of these issues have been taken into account. I refer to the evidence from the Data Protection Commission last week. The witness phrased it as "not per se" and that it would be possible if one put all of the appropriate safeguards, data protection impact assessments and all of the requirements around those provisions. That is fine but what we are saying is that there would seem to be a more straightforward way that would not figure the privacy concerns on that particular point in terms of the address.

Overall, the first point is to make sure the office of the regulator is appropriately staffed and resourced. I think that is very important and should not be lost sight of. It is possible to set down in legislation that the role of the regulator in this space will be to ensure transparency in online political advertising and not, as I said previously, go into the level of detail that is currently provided for in the general scheme. For example, one could have a statutory code, in that once there has been engagement with online platforms, academics, etc., the electoral commission would codify it in statutory code, which sets down the requirements that online platforms should meet. That would allow the online platforms to build, and to continue to build and innovate, the products and services that we have, and that are different, right across our own platforms and the other platforms that would be involved in the space.

In terms of the point made about what the European Commission is doing, it may be that what is announced by the European Commission in six months is not satisfactory to the legislators here or to Government. That is okay as Ireland, as a member state, has the power and competence to legislate for electoral law. Is it just a case of asking, given that that is only six months away, would it be more prudent to wait to see what is in that legislation before progressing with these measures as they are? There are a number of things in there. Ultimately, what we want in this space is an effective regulator to engage with and to continue to innovate and improve our services and provide transparency. That is the objective here.

I thank all of our guests for being with us today. I know that it is not easy, especially when doing things virtually but such is the world. The witnesses are probably more used to interacting online than we are.

I share the concerns as expressed by most members and witnesses here today about the online political advertising element of the electoral reform Bill being just limited to the election period. The media commission and EU laws will regulator year-round content. I also appreciate that, from the perspective of the witnesses, it would be an awful lot easier if there was seamless alignment between what we propose here, through the electoral reform Bill, and the media commission and EU regulations. From our perspective, what we are tasked with is what is in the interest of the common good and not what is necessarily easiest.

As both Facebook and Twitter have their European headquarters in Dublin, there is a huge opportunity for us to be the leaders on this issue. I wish to acknowledge the leadership shown by Facebook when it comes to transparency through its own public ad library.

I have a question for both of the witnesses. From a Twitter perspective, I ask Mr. Costello to talk us through Twitter's engagement with the EU electoral commission during its election cycle and perhaps, in part, any learnings that he thinks Ireland could benefit from.

I seek an assurance from Facebook that it will be able to respond to this legislation from a workforce management perspective. Last week, the vice-president of integrity at Facebook said on his blog:

To address this challenge, we’ve built a global network of more than 80 independent fact-checkers ... When they rate something as false, we reduce its distribution so fewer people see it.

That is 80 out of 52,000 employees worldwide or 0.15% of the workforce who are currently tasked with fact checking. I am quite concerned that this law will place additional responsibility on Facebook to monitor political advertising. That is money being spent by political parties and politicians to shape and influence democracy so this is extremely serious. I seek an assurance from Facebook that it will be able to respond appropriately and that it will be able to flex its workforce management plans so that it not just reduces content that is illegal but removes it.

I would like to pick up on a question asked by my colleague, Senator John Cummins. Perhaps Mr. Ó Broin might shed some light on the matter. The Senator asked about Facebook page managers. The example was given of a page with 11 Irish managers, seven managers from the UK, two from Serbia and one from Germany. It would be helpful if Mr. Ó Broin indicated what Facebook managers can do with political pages. Can they administrate discussions on the forum? Can they post on behalf of people or are they in control of advertising?

Mr. Ronan Costello

I might come in on the first question and then on how we worked with the Parliament, as an institution, prior to European elections. The elections in 2019 were in late May and we started talking to the Parliament team in September of the previous year. The team is apolitical. It is tasked with ensuring that the elections go off without a hitch, that the process is constructive, and that the dialogue is as informative as it can be. The benefit is that it is one project team speaking to another project team. Each of us has a single point of contact. We both know who to contact so there is an escalation route. In September, we began a conversation. Around November, a cohort from Dublin, including myself, went to Brussels to join our colleagues who work in Brussels full-time, and we gave a training session to all available MEPs and their staff on Twitter safety, trust and safety expectations that they could have, and our rules, policies and escalation routes, etc.

That work continued from there, right through from November 2018 until the election concluded on 26 May 2019. As I said, the benefit of that throughout was that there was no ambiguity as to who to reach out to in the event of X. That would be the constructive advantage of an electoral commission in Ireland.

Mr. Dualta Ó Broin

I will come in now. I thank the Deputy for her questions. We agree that Ireland has an opportunity to show global leadership in effective regulation, not just of online political advertising, but of a whole range of other issues. Harmful online content is also central to that.

I want to make a few points. Our strategy on misinformation is threefold: remove, reduce and inform. What the Deputy has referred to concerns the reduce piece. Those 80 she referred to are not Facebook employees. They are 80 employees of third-party organisations that we work with to whom we refer content that has been flagged as possibly being false. Those organisations carry out checks on whether the content is true. In Ireland, Journal.ie is the third-party fact checker. That is where that figure of 80 comes from.

Ensuring the integrity of democratic processes and elections involves a much broader group of people right across the company. I believe there are around 40 teams involved in this work. To give the committee a sense of the amount of content we remove, between October and December 2020, we removed 12 million pieces of content from Facebook that were misinformation which could lead to physical harm. In that same period, of the content, 167 million pieces of content that was rated as false by fact checkers was reduced in its circulation.

On the issue of page administration, the central point is the person who is running ads in Ireland must be in Ireland. There are other activities such as moderating discussions and using the safety tools we put in place for page administrators to moderate discussions on pages themselves. The Deputy also mentioned commenting. That is all possible as a page administrator.

I will direct my questions to Mr. Ó Broin. It is important for committee members to be very clear. The Digital Services Act does not deal with elections. Elections are the exclusive preserve of member states. Therefore, if I understand it correctly, the Digital Services Act provides the basic set of rules that should operate around digital transparency and accountability across the EU. However, because each electoral process is different, every member state is absolutely within its rights if it wants to set a different, higher or specific set of requirements during elections. I agree with Deputies Higgins and McAuliffe. I would much prefer it if it was more than just the four weeks when it comes to the financing. We agree on that.

Rather than, as Mr. Ó Broin's statement disingenuously suggests, people having to choose between violating Irish law or violating EU law, because I would hope Facebook never chooses to violate any law, there would be compliance with both sets of laws - the general Digital Services Act as it applies in ordinary times and, in any individual member state, any specific additional requirements the legislators in that jurisdiction feel are necessary to protect the integrity of their state. Just so we are clear, it is important for people to understand that the Digital Services Act does not actually deal with elections.

On the issue of microtargeting, the issue is not geographical targeting, just to go back to Deputy Cian O'Callaghan's point. It is data mining and the use of it without the knowledge of the people who are being targeted with the advertising. For example, when I get microtargeted by Amazon, it is on the basis of the data it has mined from whatever books I have viewed and browsed on Amazon. I have never knowingly consented to that data mining or indeed Facebook's use of that. Obviously, when I signed the 75-page consent document when I signed up to Facebook, there was implied consent, but there are all sorts of problems with that. That is why, for example, the European Data Protection Supervisor has argued for action on this and therefore greater scrutiny and possible banning. I would be interested in Mr. Ó Broin's view on that bit of it because I do not think anybody is talking about the geographical targeting of ads.

In Mr. Ó Broin's comments on page three of annexe 8, my reading of what he is proposing in terms of the suggested change to head 122 is that Facebook would have no responsibility for ensuring those who are advertising during an election, and from which Facebook is profiting, are legally compliant. All it would have is the responsibility, through a point of contact, to share information with the regulator and its job would be to enforce compliance. Will Mr. Ó Broin confirm that that is his position.

Some important points were made in our previous session about changes to the way in which online operators operate and the kinds of tools that are at their disposal. There was an interesting article in The New York Times in December of last year that mentioned Facebook withdrawing what was referred to as its "news ecosystem quality", or NEQ, which is an algorithm for prioritising more informed and reliable sources rather than less informed and reliable sources, which was in train during the election and withdrawn afterwards. One of the big issues here is how we all will know if new tools are introduced or withdrawn and how they are captured. They are also particularly relevant to electoral integrity. What transparency mechanisms does Mr. Ó Broin propose in respect of that?

Mr. Dualta Ó Broin

I will take each of the issues in turn. The Deputy is correct to state that the Digital Services Act, DSA, does not, as far as I know, seek to provide for transparency in political advertising. What it will do is provide an EU-wide baseline for all advertising. The European Commission has also said, however, that through the European democracy action plan, EDAP, it intends to introduce a separate legislative instrument which will provide for transparency in political advertising on an EU-wide basis. We expect that to appear in the next six months. The DSA provides the baseline for all advertising, but the EDAP will be a legislative instrument which introduces requirements for transparency in political advertising across the EU.

I accept the point that electoral law is a member state competence and the organising of elections falls under electoral law. It would seem to us, however, given this instrument is just on the horizon, that we should at least wait to see what it is or what its scope is going to be before bringing in these detailed provisions at this time. I do not want to suggest to the committee and I know it has spent a long time on this already, but it may be helpful - I do not know - to hear from the European Commission as to what its plans are in this space.

On responsibility and compliance, there is no issue with taking responsibility for getting details from an individual and putting them in the ads library. It is once it gets beyond that into the sources of funds and all of the other matters which are considered beyond that that we will take information from the individual who is looking to run the ad and, if there are investigations etc. beyond that, that we see it as a role for the regulator to play.

On the point of tools and changes, we make announcements where there are changes to any of our tools which require notifications to the appropriate regulator Obviously, we engage with the Data Protection Commission on an ongoing basis in respect of any changes we make to products and services, as appropriate in the EU.

Returning to the second question, if I understand it correctly, the purpose of head 122 is to place some level of responsibility on the platform to ensure it is not accepting, and therefore profiting from, advertisements that are not compliant with the law. Is Mr. Ó Broin saying the platform should or should not have that responsibility? The regulator obviously has its parallel responsibility, but should there not be some responsibility on Facebook to endeavour to ensure that an advertisement is legally compliant with the provisions of the Act?

Mr. Dualta Ó Broin

I think we are coming at this issue from slightly different perspectives. I am saying there is a more effective and efficient way of doing that to ensure we take on our correct responsibility. However, we also see that there is a crucial role for the regulator to play here rather than this third party model.

Just to be clear, Mr. Ó Broin is saying Facebook's sole responsibility would be to furnish information to the regulator, which would then have sole responsibility to ensure compliance. Facebook would have no responsibility to ensure compliance in the case of, for example, a politician placing a non-compliant advertisement on its platform.

Mr. Dualta Ó Broin

No, that is not what I am saying. If the law or the code requires us to check the identification or passport of an individual, we are obviously responsible for doing that. It is the model of oversight that I am trying to make a point about. I am not sure this third party model is fit for the online world. It would seem far more appropriate to set down what our roles, responsibilities and obligations are under the law and that this third party would be replaced by more direct involvement by the regulator.

I thank Mr. Ó Broin.

I thank the witnesses for appearing before the committee and helping us to complete this pre-legislative scrutiny. My first question is for Mr. Costello. I appreciate that Twitter's has fewer users in Ireland than Facebook. I understand the figure is less than 1 million people. I also understand that in recent times Twitter has taken more direct action to remove bots and trolls. How many accounts is Twitter removing in Ireland every month?

Mr. Ronan Costello

The short answer is that I do not have a figure for Ireland specifically.

A global number would be fine if an Irish one is not available.

Mr. Ronan Costello

At the Twitter transparency centre, available at https://transparency.twitter.com, about every six months or so we update that site with the latest figures on our enforcement. That would include all violations of Twitter rules and the information is broken down by rules. If we were talking about platform manipulation and spam, for example, which would be co-ordinated inauthentic activity, bots if one likes, those kinds of figures could be found there every six months, depending on the nature of the report.

This figure is not up to date but as an example, around 18 or 24 months ago, we were challenging between 8 million and 10 million accounts globally each week. Challenging means that something about the sign-up process with particular accounts was setting off flags. I mentioned earlier that we are leveraging machine learning to be better at detecting behaviour-based signals. Machine learning flags were tripped and an account would be challenged. It is based on the suspicion that it is an inauthentic or spam account, which is likely therefore to engage in inauthentic behaviour on the platform, perhaps in co-ordination with other accounts and-or as part of a network. If the challenge is not met by an account and it does not prove that it is being operated by a real individual and that person has not had accounts suspended previously, the account in question will be automatically suspended. There was a significant volume of such actions globally around 18 or 24 months ago. There is still a significant volume weekly but it is not as high as it was. I suggest that is evidence of our enhanced capacity to deploy those machine learning techniques to address this sort of behaviour.

The response to my next question varies a little depending on the way the question is asked. To put it simply, does Facebook fact-check political advertisements?

Mr. Dualta Ó Broin

No. We have taken the position that we do not fact-check political speech, whether it is in organic content or political advertisements. We believe it is right that political speech should be seen and should be scrutinised more than it already is by placing it in an advertisements library. It is important to make the distinction, however, that this does not mean anything can be said by a politician in a political advertisement. Hate speech, for example, is not allowed, nor is suppression of voting, election turnout or that type of behaviour. A whole range of content is not allowed. As a general point, we do not fact-check and we do not send political speech material to the network of fact-checking organisations I spoke of earlier.

I am familiar with Facebook's community guidelines. If Facebook does not fact-check, that means it does not check whether an advertisement is true.

Mr. Dualta Ó Broin

We do not believe we should be deciding what is true, which is why we work with the network of third party fact-checkers. For non-political speech, they decide whether content concerning Covid-----

Who are these third parties that have more authority than Facebook to determine whether something is true?

Mr. Dualta Ó Broin

They are members of the international network of fact-checkers. In Ireland, the organisation we work with is TheJournal.ie.

TheJournal.ie does the fact-checking for Facebook on all Irish political advertisements.

Mr. Dualta Ó Broin

Is the Senator referring to political speech?

I am specifically questioning the veracity of political advertisements on Facebook in Ireland.

Mr. Dualta Ó Broin

The traditional approach for organic content in respect of determining whether it is true is that it is sent to a third party fact-checker. We do not do that for political speech in advertisements or organic content. It is not fact-checked. It is allowed in the business but there are barriers set down in the community standards regarding what is allowable.

Why is truth not a corporate value for Facebook?

Mr. Dualta Ó Broin

We do not believe it should be for companies such as ours to make the decision on where the line of truth is. In political debate and discourse there can naturally be different perspectives on certain things. We are, however, committed to external oversight to ensure our standards and policies are exactly where they need to be. That is, for example, why the external oversight board was established. It oversees the decisions made and it is the final arbiter of the decisions we are making regarding content. The board is also going into other areas and recommendations, and we are engaging with it on those. It comes down ultimately to the question of who should be the arbiter of what the truth is when it comes to political speech.

I thank Mr. Ó Broin.

As this is the last round of questions for me, I am somewhat concerned by the responses to Senator Fitzpatrick's forensic questioning. That is for another day, however. I am going away now more uncomfortable than I was when I came in. I say that particularly regarding Facebook and its representative. I have two simple questions. Whenever we talk about a fact-checking or watchdog-type body, individual or team, we must always ask who will watch the watchmen. That is a profound question. Who will watch the watchman? What mechanisms do Twitter or Facebook have in place for oversight and accountability?

That is something about which we need to hear. Have the witnesses individually, or the organisations they represent, in this case Facebook and Twitter, personally, professionally or as a collective in terms of those organisations, been involved in any lobbying in regard to this legislation? If so, with whom? They might share that with the committee. I ask that question because it is important. There is nothing wrong with lobbying and it is a regulated profession. However, it is important, as a committee doing pre-legislative scrutiny in regard to this legislation, that we should know. I would like to hear what views the witnesses have and what their engagement has been with lobbying the political institutions, in particular politicians in Leinster House. I want to be specific about my questioning because I have thought about it and I have knowledge about it. Specifically in regard to this Bill, can the witnesses share, either as individuals or on behalf of the organisations they represent, what lobbying they have been involved in?

Mr. Ronan Costello

The point on oversight and accountability is an important one. For example, with regard to our innovations around labelling of harmful and misleading information, one of our motivations for doing that is to give primacy to, and to promote content from, authoritative sources. For example, when we label content around elections and we say that this claim about the outcome of an election is disputed, we are basing that on the outcomes and the statements of the electoral commissions. In the recent presidential election in the United States, for example, it could be a state or county board, or something like that. One of the reasons we are doing that is not only to make the users’ experience on Twitter better or more reliable, so they are getting good quality information rather than poor quality information, but also because we would hope that it buttresses and reaffirms faith in those institutions that have oversight of these processes. We are relying on their determinations and want to give them primacy on our platform.

With respect to our own oversight and accountability, I will give an example. The policies we develop, be they around misleading information, abuse, harassment or hateful conduct, are developed in collaboration with our global trust and safety council, which is a collection of non-profit organisations, typically NGOs, expert groups and sometimes individuals who are experts in their respective areas. They say to us, for example, that we do not have a policy on something and we should have a policy on it, or that our policy on it is lacking and it is not taking account of trends in this particular area, and that we need to do some development work on it. That is an external body. All of those engage with us independently and their feedback is very candid.

What about lobbying? I address the question to both individuals.

Mr. Ronan Costello

This is my first engagement on this. The only written submission we have made with respect to this legislation is the one the Senator has before him.

I ask the same question of Facebook.

Mr. Dualta Ó Broin

On the point on oversight and who will be watching the watchmen, we have the oversight board in place, which we established, which is independent of us and which is a global board that will be the final arbiter on whether we made the right or the wrong decision in removing individual pieces of content. In addition, we realise that external oversight of our systems is incredibly important in establishing and building trust. This year, an independent third party has responded to a request for tender to carry out an audit of our systems and services right across the world, the systems and services we have in place to protect and keep users as safe as possible on our platforms. It is the first time it is being done. That might feed into the approach that is going to be taken under the digital services Act to establish whether, at the systems level, the company is doing what it should do.

In respect of lobbying in regard to this legislation, the first we heard of it was in The Irish Times article in early January. We held a meeting with officials from the Department of Housing, Local Government and Heritage. We had not established our position at that time and wanted to understand what was the timeframe. We have since put together the written submission of 23 February, which all committee members have. We sought to engage with members of the committee and to have individual meetings with members of the committee in the run-up to this. We were not aware this hearing would happen but, obviously, the committee has our opening statements and all of the details. That will all be reflected in the lobbying register the next time the reporting is due.

We have had a good discussion around paid political advertising. I have one remaining question. European regulation of online political advertising would provide harmonisation, and I can see why multinational companies would rather have that framework than a different framework in each country, for operational reasons alone. The difficulty is that different countries allow different levels of donations. For example, on this island we have two different jurisdictions, one with almost an uncapped donations system and the other, in the South, which, to be fair, bans corporate donations above €100 and caps personal donations. When one limits the amount of money in politics, the exploitation of the tools becomes less serious but when there are uncapped resources, that is a difficulty. I would argue that while European regulations are important, we need to specify for each particular country and each democracy. Our voting system is unique, with the exception of Malta.

My question is for Twitter. We have talked about paid political advertising. I believe the allowing of anonymous accounts on Twitter also prevents transparency around the spending of money in politics. Yes, that can happen on Facebook but it is far more obvious on Facebook when an account is fake. I believe the anonymous accounts on Twitter allow a disgusting and horrible culture of trolling on Twitter, and I do not understand why Twitter would allow that to continue. I have continuously called for anonymised accounts to be banned, and I do not know why Twitter allows it.

It would also prevent paid political operatives operating multiple anonymous accounts, which prevents transparency as well. If I am paying for election posters to be erected or leaflets to be delivered, people can see that. However, if I am paying people to troll other people on an account, that cannot be seen. I am not saying that is happening but that it is very possible under the current system. Anonymised accounts facilitate that, and I cannot understand why Twitter allows those anonymised accounts on its platform.

Mr. Ronan Costello

We have had a policy of allowing pseudonymity on the platform since the platform's inception. One of the reasons that has not changed is that the policy was developed and has evolved in consultation with human rights groups. For example, groups like Civil Rights Defenders, Access Now and Front Line Defenders are based in Dublin and all of them represent human rights activists across the globe. One of the reasons we allow the pseudonymous accounts, and one of the reasons I mention the human rights groups, is that that particular feature of the platform is vital in environments, contexts and countries where it is dangerous for that kind of activity to be carried under one’s real name because one can be identified and may be apprehended, and so on. We see this context and we see this use case arising more and more as the political situation in countries across the world deteriorates and regresses into a form of a illiberalism.

Pseudonymity on Twitter is one way for folks who oppose that developing illiberalism to push back on it and to do so not altogether with safety but perhaps in safer conditions than they would be able to otherwise.

It is also in a way that permits trolling.

I accept the pseudonymity argument to a certain extent but I believe social media companies need to do more in respect of the trolling of accounts. I will look at the link that Mr. Costello referred me to by way of an answer to my first question, but I am concerned that it will not give me the answer I am looking for. I will come back to Mr. Costello, if that is okay, because I would like specific answers on that.

My thanks to the Facebook representative for participating today and for making the submissions. We need to recognise the big divergence between how Facebook is approaching this issue and how we are approaching it in terms of trying to protect the fairness and freedom of our electoral process. Facebook has 2.4 million users in Ireland and the company is saying there is no fact checking. The submission suggested Facebook did not want any responsibility in terms of compliance with advertising and regulatory standards for political advertising. Facebook has the tools to demonstrate and contribute to making our democracy stronger. I urge Facebook to work with us in the coming months to try to ensure that we can get the maximum out of the tools it has to enhance our democracy. I will leave it at that.

My thanks to the company representatives for being here today and for the answers we have received. Last week we heard a good deal about the precautionary principle. When it comes to this, I believe the horse has bolted with regard to the use of social media. Advertising is no longer what influences political discourse. Rather, it is the posts and the use of social media. Where we increasingly regulate paid advertising, what we will see is an increase in the non-paid activities and activities that are encouraged by the platforms represented here today and other platforms. The more outrageous content is and the more eyeballs the companies get on screen, the easier it is for them to make sellable advertising for micro-targeting, even outside of the political arena. There is a business model at the back of all of this. We have heard all the platitudes and the niceties about political discourse but behind all of this is a business model that drives as many eyeballs on screen as possible to sell advertising. I believe that this section of the Bill is fiddling while Rome burns, to be honest.

Professor Nathanial Persily of Standard University has written extensively on the idea of the dangers posed by social media and the things valued by the companies. Velocity, that is to say, the speed of communication is one. This has a tension with democracy. A well-timed lie immediately in advance of an election can gain a national audience long before it can have the opportunity to be rebutted.

The second is the idea of things going viral. The primacy of what goes viral is what is emotive and outrageous. Consequently, democratic discourse is not well served where there is a chase for the outrageous. The outrageous is less well regulated, and positively encouraged in some instances by suggesting that people join other groups or join groups of similar interest. We have incentivised eyeballs on screen and what we have overall is a coarsening of public discourse as a consequence.

Anonymity is another point. It is some years since I set up my Twitter account so I had forgotten the process. Last night, for the craic and to see how easy it was, I set up a Twitter account under a different name. All that was necessary was a mobile telephone. In doing that there is no accountability in place. There is no obligation for me to identify who I am. There is an opportunity to spread lies, misinformation and disinformation from within and without the jurisdiction and, consequently, influence political discourse in advance of an election. The professor goes on to speak about the monopoly held by platforms and the challenge to the circumstances where social media posts are made worldwide. We have seen that in social media being managed from other countries. That is a problem for sovereignty and a problem of oversight and of governance. In light of this, Twitter says it is encouraging civic participation and dialogue. Facebook is a flag-waver for free speech. To be fair, Facebook is taking positive steps. However, cyber-scientists have shown maps of trolling and pile-on activity and have confirmed that there are instances where this is co-ordinated or orchestrated for political purposes with a view to suppressing content, silencing people and influencing discourse. Both platforms represented before the committee promote that or sit by and allow it to happen. How do the companies expect us to regulate that? What are the comments of the company representatives in response to that? The provision in this Bill is absolutely not fit for purpose. It is too narrow and does not protect our democracy.

Mr. Ronan Costello

I might come in on that, especially because it gives me an opportunity to say something around pseudonymity that I did not say in my first answer. It is the case that a person can have a pseudonymous account. A user need not use a real name on Twitter for the reasons I outlined in my first answer. The second point is that, regardless of the name or handle of the account, the enforcement actions that we take against accounts is the same no matter what. If a person is abusing or harassing someone under a pseudonym or engaging in a spam operation where a person is trying to develop momentum around a particular hashtag or topic such that it gains inorganic primacy on the platform, then we will enforce our rules in the same way against that pseudonymous activity as we would if the person was operating under a real name. The enforcement outcome is the same no matter what. I hope that it clarifies our approach on this.

I am sorry to cut across Mr. Costello but that does not answer it. The fact is that anonymous parties with no accountability or transparency in respect of who they are, their political leaning, where they are or otherwise get to influence the discourse. We can look at anything that is trending in Ireland on any evening. If we follow in through it, there are accounts and the promoting of particular hashtags that put out information that is false, that misrepresent how people are elected in Ireland and what is valued in people being elected or that misrepresent the proportional representation system. There are people behind this who do not put up their names and who promote misinformation and disinformation about Irish politicians.

Mr. Ronan Costello

My response is that they will receive the same treatment with respect to enforcement of our rules as an account user using a real name.

Except that it has gone viral by the time the company does that. The information is out and around the country by the time Twitter does that so it is too late to be rebutted.

Mr. Ronan Costello

There are two things I would say to that. One of the benefits of Twitter being a public platform is that rebuttal can be done as publicly as the sharing of disinformation. Someone can quote-Tweet in respect of a piece of disinformation that the person believes to be disinformation and say it is entirely wrong and include a link to the correct source, whether the Electoral Commission or a news organisation like Reuters or AP. If that rebuttal is done using an account of a journalist who is verified and who has 20,000 followers, it will have a larger impact on correcting the record than an pseudonymous account that is part of a low-scale co-ordination network that may only have 50 or 100 followers.

I have been handed a Tweet by, Aoife Moore, my colleague. It makes reference to watching social media companies talk about their content management and how they are so quick at removing posts while the person is constantly told that an anonymous troll account, which routinely calls the person a terrorist, has not broken any rules. That same experience has been encountered by journalists and female politicians in particular. I have had queries come back to me stating that the company will not remove the content or take action because it does not correspond with the values of the company. All the while the misrepresenting of democracy in Ireland is continuing on the platform and the coarsening of public discourse is being facilitated.

I am sorry, but I must move on as we are tight on time. We are proceeding with the third round. Senator McGahon will be followed by Senator Cummins and Deputy Ó Broin.

Thank you, Chairman, for allowing me to attend and join in, as I am not a member of the committee. The reason I am here today is that I have taken a great interest in transparency on these platforms since my election to the Oireachtas. I wish to follow up on what my colleague, Senator Seery Kearney, said. With all due respect to Mr. Costello, under no circumstances do I accept the accept the excuse that Twitter is somehow liberating oppressed people in different parts of the world and that is why it is necessary to have anonymous accounts. That is total and utter BS. It is a nice, fluffy answer. Instead, the anonymous accounts are dishing out some of the most horrific abuse on the platform. Every time I pick up my telephone and see Twitter notifications I get pangs of fright and fear about what is next. It is the same for every politician. I would not be on Twitter if I did not need to be for work and politics reasons. It has reached the stage where I believe it is beginning to undermine our democracy.

Twitter is 15 years old and I do not buy the case that this is how Twitter started 15 years ago. Political discourse on Twitter has changed. Ten years ago, Twitter was a wonderful platform where one could engage with people who had different viewpoints. One could engage in and have respectful debates. That is gone out the window now. It makes zero sense to have anonymous accounts that can be set up with five different emails for five different accounts to target politicians, journalists such as Aoife Moore or members of civic society. That is how pile-ons are orchestrated. The point is that there are only so many times one can be called every word under the sun before it rolls off one's back. It is different for some of us because we are used to it. However, if social media companies are allowed to continue with this type of political discourse, we are going to lose some potentially really good young people from politics in Ireland. An amount of people ask me why I even got into politics given the abuse I receive online. They say they would run a million miles from it.

That is the problem with this, and it is why I believe platforms such as Twitter and Facebook are undermining democracies. I do not buy their reasoning that they maintain anonymous accounts because they can unveil corrupt places in far-flung parts of the world. I am referring to anonymous accounts in this country.

As regards Facebook, my colleague, Senator Seery Kearney, is quite correct. By the time disinformation spreads on Facebook, the horse has bolted and it is too late to rebut it. If one tries to rebut something, there is the Streisand effect whereby one is making a bigger story of it by trying to rebut it. When one of us rebuts something that could be completely false, that is then in the newspaper. It is the Streisand effect, so it is not in our interests to try to rebut fake news about us as individuals or the like.

The other point is that there has been much discussion today about what we are doing to fact-check politicians and political parties, but what are we doing to try to fact-check the public? What is becoming a thing now in elections is that in the seven days before an election, which could be a crucial or tight election, people are going on Facebook, Twitter and other online sites to undermine candidates. The new form of online political campaigning is to try to smear candidates. A former Deputy is now suing Facebook and Twitter for damages because of completely unfounded lies that were put forward about him in the lead up to the last general election. Fair play to him for taking it to the High Court. I commend him on that. The point is, however, that the damage was done. The issue is that people see it online and believe it, due to the age-old concept that there is no smoke without fire. There should be a direct link whereby somebody can go to Facebook. The person should not have to go through a lengthy court process, only for somebody to say that he or she put it up as a lie and did not mean it.

Senator McGahon, if you have a question, there will be one minute left for a reply.

What are the witnesses doing to reduce the horrific abuse that is doled out on their platforms to politicians, journalists and members of civic society? They have said that they regulate anonymous accounts in the same way as they do other accounts. It is far too slow. The platforms have failed to regulate themselves and now governments need to do it.

Mr. Costello or Mr. Ó Broin have 30 seconds to answer that question.

Mr. Dualta Ó Broin

I am happy to reply.

Go ahead.

Mr. Dualta Ó Broin

On the point about the safety of politicians, we recognise that this is an issue of concern at societal level. That is why in advance of last year's general election we held a training session that was open to all candidates on the safety and security features that are in place on the platform. We followed that up for all successful candidates afterwards. We have put in place a dedicated email alias, which is OirSafety@fb.com, for politicians to report issues of concern about their safety to us directly.

Thank you.

Mr. Dualta Ó Broin

It is not just that. We are also evolving the policies in that space and engaging with Irish stakeholder bodies on where the policy relating to public figures should be.

Senator Cummins has three minutes and then Deputy Ó Broin has two minutes.

I will only take two minutes.

Just direct questions and answers, please.

I echo many of the comments that have been made about anonymous accounts. It is unacceptable in this day and age. If online banking apps have the capacity to be able to verify identities, there is no reason that online platforms cannot do it. To return to my original question to Facebook relating to advertising, what I am concerned about is outside interference in our democracy. I have been on the platform over the last hour while engaging in this meeting and one can change the attribution setting during campaign creation. One can change it from advertiser to administrator to analyst. Is that correct?

Mr. Dualta Ó Broin

The key point is that the person who is involved in the political advertising has to be in Ireland. That is the requirement.

However, one can change the attribution setting during campaign creation. If my Ireland manager sets up a draft campaign, is it possible for my manager in Serbia or Germany to go in, add to that campaign, create data sets and then for another advertising manager in Ireland to press the "run" button on it? With all due respect, it is a "Yes" or "No" answer.

Mr. Dualta Ó Broin

What I can say is that the requirements are that anybody involved in-----

Requirements are different from a "Yes" or "No" answer. Is it possible for my Ireland manager to set up a draft campaign, entitled "Public Meeting", and for one of my managers in another country, with no connection to this jurisdiction, to add demographics or create further content within that campaign and then for another Ireland manager to press the "run" button, "Yes" or "No"?

Mr. Dualta Ó Broin

It is my understanding that it is not possible.

It is Mr. Ó Broin's understanding. It would be very important for this committee to have a "Yes" or "No" answer to that question because my understanding from my analysis of the platform is that it is possible.

Mr. Dualta Ó Broin

I am happy to follow up.

I thank Mr. Ó Broin.

We could possibly get a written answer to that question. If Senator Cummins would outline that in a written question and we get a response, it would be of assistance to us. I call Deputy Ó Broin.

Absolutely. I will be brief because Deputy Gould has a question.

We have two minutes left.

Yes, or less than that. I agree with Senator Seery Kearney, but the responsibility to legislate is here. If people want, for example, anonymised accounts or micro targeting ended and better regulation of these platforms, we legislate for it here. There is no use in Government Deputies coming here and complaining that there have been 15 years of failure in self-regulation by the industry when we know self-regulation in many other areas has failed. The challenge I put to the Deputies and Senators, and I accept their sincerity because I share many of their concerns, is to regulate. If they will not do that, people will question their sincerity when they come here to rail at platforms for not regulating when that is our job. Beyond this Bill, which we all have accepted is very limited in its regulation of online advertising, there is clearly a bigger job of work for this House to do. I believe members would have willing partners across the Chambers to regulate platforms properly to deal with all these issues. Let us go and do it. Then we will not be asking industry to regulate itself because we will have taken our responsibilities as legislators seriously.

The EU cannot deal with the transfer of data out of the EU to the US by Facebook. I am finding it very difficult. A little country like Ireland on its own cannot do that-----

With all due respect-----



Bigger issues cannot be dealt with------

Under section 140 of the Internet Safety Act-----


There is the opportunity-----


I remind Senator McGahon to speak through the Chair. Thank you. I call on Deputy Gould to ask a direct question, because we are out of time.

For most people, social media can be entertaining and is an enjoyable way to interact with friends, acquaintances and the community. There are great positives with social media, but what has happened over time is that social media has been hijacked by bullies and organisations with hidden agendas and that want to attack people. A party colleague of mine has been attacked constantly by one person on Facebook over a period of years. She has not been able to do anything to stop these attacks. As others have said here today, they are not the only people affected by this.

People should have the right to freedom of expression but they cannot be allowed to bully, harass and attack people. Facebook must use its moral compass to do something to protect people. Anonymous accounts are unregulated. Facebook and Twitter are huge companies. We have seen what has happened in the US. Surely, they have learned lessons that they can put in place in a country like Ireland, where we have a much stronger democracy and people have the right to be protected.

Mr. Dualta Ó Broin

I can jump in there, briefly.

I ask Mr. Ó Broin to be as brief as possible; we are out of time.

Mr. Dualta Ó Broin

We welcome regulation in respect of harmful content and political advertising, etc. One thing that can be done is to expedite the establishment of an online safety commissioner and to implement the EU-wide rules on the audiovisual media services directive, which is being dealt with by the Media Commission at the moment.

I thank the witnesses for their attendance today; it has been most helpful.

The joint committee adjourned at 11.33 p.m. until 12.30 p.m. on Tuesday, 6 April 2021.