Dáil Éireann Debate, Thursday - 29 June 2023
15. Deputy Paul McAuliffe asked the Minister for Enterprise, Trade and Employment if he agrees that without appropriate regulation, the increased usage of artificial intelligence, AI, technologies poses potential risks to workers; and if he will make a statement on the matter. [31386/23]
Amharc ar fhreagraWhat steps or proposals, if any, does the Government have to regulate AI? I know the State generally does not stick its nose into technological advancements. However, because this is such a precarious and dangerous form of technology that needs to be carefully managed, I would like to know whether there are proposals within the EU or this State to regulate it.
The draft EU AI Act will set out harmonised rules for the design, development, placement on the market and use of AI systems in the Union. The Act will ensure the protection of fundamental rights and user safety as well as build trust in the development and uptake of AI, thereby enhancing EU investment and innovation. The proposal aims to address the risks generated by specific uses of AI through a set of complementary, proportionate and flexible rules. Ireland very much welcomes the development of the AI Act and the progress that is being made at Council and at the European Parliament to advance this important Act. Given that AI has evolved significantly in the past few months, it is important that this regulation is flexible and future-proofed in order to ensure that it continues to protect the safety and fundamental rights of the individual while also ensuring that innovation for good continues in this area.
The European Parliament agreed its approach by plenary vote on 14 June. Officials in my Department are examining the text and consulting with relevant stakeholders to refine Ireland's position on the agreed text. We are supportive of measures that provide clarity and reassurance to all stakeholders engaged in this complex and evolving area, particularly measures that help ensure a level playing field in Europe. We are aware that there are still matters that require more detailed discussion and this is a priority for my Department in the coming weeks and months. It is already clear from the first trilogue meeting that some issues, such as an agreed definition of AI, will be more difficult to resolve than others. However, given how quickly this technology is evolving, it is clear that there is a willingness among all parties to move forward and reach agreement on this regulation by the end of 2023 or very early in 2024.
Additional information not given on the floor of the House
Final agreement on the Act is expected by late 2023 or very early into 2024. Existing law also applies in the context of AI, so all of the legal protections for workers are still available but the EU AI Act will add an additional layer of protection where AI tools are deployed.
The national AI strategy, AI - Here for Good, was launched in July 2021 and set out a vision that Ireland would be an international leader in using AI to the benefit of our population through a people-centred and ethical approach to AI development, adoption and use. Much progress has been made since then in preparing for AI, including by putting appropriate safeguards in place.
Of course, as with previous technological advances, changes to the world of work are to be expected but it is likely that much of the disruption caused by AI will result in changes to job roles, tasks and distribution rather than actual job losses. Indeed, a recent March 2023 OECD study on the impact of AI on the workplace found that AI users were more than four times as likely to say that AI had improved working conditions as to say that AI had worsened them. The report found that while AI has automated many repetitive and dangerous tasks, it has also created new tasks. The key recommendation of the report is that the adoption of Al results in significant skill changes - not just specialised AI skills but human skills such as creativity and communication.
Last year, at the request of my Department, the expert group on future skills needs published AI Skills: A Preliminary assessment of the skills needed for the deployment, management and regulation of artificial intelligence, a review of the skills implications of AI over the next five to ten years, as well as the skills-related actions needed to realise the potential of AI. My officials are working closely with the Department of Further and Higher Education, Research, Innovation and Science to ensure the recommendations of the review are acted on.
I welcome the fact that there is such work being done in the Department on this issue. It is appropriate that it is being regulated on a European-wide basis as it would be ineffective if we were just dealing with it domestically.
As the Minister of State is aware, technology has had significant benefits for this country and the global economy and nobody is suggesting that we should stop the march of technology but when you look at the potential consequences of AI, it is important that we recognise that this is a form of technology that we cannot allow to just develop the way the market desires it to. Obviously, it will have enormous consequences in the employment sphere. I suspect very many jobs we have taken for granted may be eliminated as a result of it. That may be difficult to regulate but what we certainly have to regulate is the impact it may have on democracy in terms of trying to present images as true when they are false. Will issues like that be considered in the context of the EU AI Act? Is this an issue that deserves to be considered?
We had a very good session on AI at the Joint Committee on Enterprise, Trade and Employment last week. Deputy Jim O'Callaghan raised the issue of jobs being displaced. I would have a particular concern about jobs being displaced by AI. Effectively, we are talking here about a platform worker who has not necessarily been sacked but who, because of the way the algorithm works, does not get any shifts the following week. If you are not working, it is the same as being sacked. There would be no involvement of any individual to take into account the worker's circumstances in this case. There are all sorts of reasons why a worker might not turn up for work - it is easily explicable - but if we are only dealing with an algorithm, it is a case of ticking that box and the person does not get any shifts the following week. At any stage in this process where there is an impact on workers, there must be human involvement. This was something the committee discussed. This is moving very quickly. I have a concern that some of the processes are moving far slower than this technology.
As I said to Deputy O’Reilly yesterday, I invite all Deputies and Senators to come to the audiovisual room on 11 July, when Dr. Patricia Scanlon, the national AI ambassador, will answer many of these questions. Regarding the legislation, we are looking at issues like the definition of AI, enforcement and governance, biometric identification and the various foundation models. We want to regulate the threats of AI such as the threat to employment and threats to people generally but, equally, we must recognise that there are enormous opportunities here. Technologies that have become common in our daily lives were also considered threats that could bring about mass extinction of employment, but we evolved, dealt with them and put appropriate guideline regulations in place that allowed for innovation and that protected citizens, not just in the area of employment but also in other areas. What we are trying to do as a country is to get a balance. My predecessor, Deputy Troy, published our national AI strategy in 2021, which was way before this debate got to the level it has. We are ahead of the game on this and we will continue, as a country, to stay ahead of the game.
To follow on from what Deputy O’Reilly said about the threat to employment, it could have vast consequences for all types of employment. For example, in the case of lawyers who draft contracts, this could be done completely by AI in the future. People may think that it is a good development. However, it will have societal consequences we need to keep on top of. My concern is that it will result globally in the concentration of wealth in a smaller group of people as the necessity for labour to create capital is broken.
We also need to look at the consequences for democracy and the world of politics. One of the consequences of AI will be that we simply will not be able to trust or believe anything we see with our own eyes. We will see images of politicians saying things they never said and doing things they never did, so we need to protect democracy when it comes to that as well. I hope this will be contemplated in the considerations of the EU and the Department.
The protection of democracy is extremely important. We have seen rogue actors trying to influence elections. There are multiple examples of that across democracies. AI is another more advanced tool for state and non-state actors to try to intervene and affect election results. My understanding is that Art O'Leary, the head of the Electoral Commission, is very engaged with this issue in terms of trying to ensure that we do everything we can to look at international best practice in order to protect against the potential threats in the context of an election cycle. The latter will obviously get under way next year in the context of the local and European elections and will lead into the next general election.
From talking to technology companies, I am aware that there is an understanding - even in the tech sector - that the threats from AI are real. They want and need a regulated environment that can allow the benefits of AI, which are significant. Ireland has great advantages from being part of the research and innovation programmes that develop AI services. Our role needs to be to facilitate appropriate regulation to protect consumers. However, we must not try to swim against the tide and allow AI technology and innovation to happen outside the European Union, which will ultimately impact us anyway. This is about trying to keep it between the ditches by having appropriate regulation while at the same time facilitating much of the global research out of Ireland.
