The European Commission’s proposals for the regulation of artificial intelligence (AI) – the AI Act – are currently being debated in the Council and the European Parliament. The Swedish MEP most up-to-speed on these issues is Jörgen Warborn (EPP/Moderate Party), so we asked him to share his thoughts on the proposals.
Increased use of AI is widely considered to be vital to competitiveness. It is also important for security, good customer experiences and the provision of efficient and effective services for citizens. To be successful this requires trust, a high degree of knowledge, the necessary infrastructure, and an enabling regulatory framework.
However, there is growing concern among businesses that the European Commission’s plans to regulate the use of AI, the so-called AI Act, is far too comprehensive and includes a large number of vague legal concepts that prevent it from encouraging greater use of AI. The proposals would be difficult to apply in practice and result in increased costs. The AI Act regulates in unnecessary detail, and overlaps and contravenes other regulations in ways that few businesses can cope with. The proposals would result in unrealistic costs and administrative burdens, not least for start-ups and small and medium-sized enterprises.
Jörgen Warborn, what do you see as the benefits of increased use of AI for society and business?
– I see enormous benefits for virtually all areas of society. We already see concrete examples such as AI-equipped cars and trucks that prevent collisions, accidents and deaths. AI also has life-saving capabilities in healthcare where it anticipates risk factors in patients and helps doctors implement the right treatments at the right time. For business, I see opportunities in self-driving vehicles, customised consumer experiences, purchase forecasting, for example, and so much more.
– The opportunities are endless. It’s also in the nature of technological development that today it’s impossible to know for sure what challenges lie ahead of us and what needs and wishes people will have in the future. It is these factors that drive the desire to innovate. Used properly, AI can enhance human performance, save money, and relieve us from repetitive tasks. When man and machine co-operate and engage in the tasks for which each of them is best-equipped, then the sum of the whole becomes greater than its individual parts. Together with new technical solutions, we can live freer, better and happier lives.
Artificial intelligence is the rocket that will take us into the future, so we need to fuel it properly.
How did you become interested in working on AI issues?
– I’ve always been a huge technology optimist. I ran IT and telecom companies myself in the 90s and I’ve always been curious about new technologies. I believe that in the not-too-distant future, AI will become as natural a part of society as electricity. And whoever wins the battle for digital technology will succeed in the future. A leading position in digital will be crucial to guarantee the prosperity of future generations. When the EU said it was going to set up a special AI committee, I thought it sounded exciting and timely.
What is your overall position on the AI Act? What are the risks and advantages of the proposals?
– In general, I’m worried about the attempt to regulate more and to regulate harder. Yes, updated and clearer regulations are needed, but a wet blanket should not be thrown over those who want to contribute to positive technological development. In public debate, there is an over-focus on the risks of AI and an overly sceptical attitude towards its potential. This creates imbalance and slows progress. The degree of regulation the proposals would entail is too cost-driving and inhibiting - and this means that the solutions that we are likely to deploy in the EU will in future be developed in other countries and we will lose prosperity growth and control over time.
– I’m not convinced that we need a technology-specific AI directive at all. The most negative aspects of AI it seeks to regulate are already regulated by other laws, so there’s a risk of over-regulation. It is already forbidden to inflict intentional harm on others. It is already forbidden to discriminate and infringe people’s privacy, etc.
– But if we are now forced to have an AI directive, then a risk-based approach is right. We must make a clear distinction between AI use that entails high risks and AI use that does not. Legislators must focus with precision on high-risk AI – not with a shotgun-approach against all AI development.
I’m not convinced that we need a technology-specific AI directive at all.
There is concern in the business world that the AI Act may be difficult to apply in practice; that it would result in increased costs and an increased regulatory burden. What’s your position on that?
– I share that concern. It is absolutely vital that the legislation is easy to implement. Otherwise, there’s a risk that development in start-ups and small businesses will be undermined. The legislation has widely differing impacts. Large companies with significant resources are able to hire consultants or have in-house experts – this is not the case for most small businesses.
– I will work to ensure that the legislation is as unbureaucratic as possible. Legislation must seriously focus on securing sensitive personal data and preventing serious risks – not complicating things unnecessarily for small businesses and organisations that do not use high-risk AI programmes. The key thing is to create predictability, with authorities helping businesses to do the right thing instead of appearing with a stick and a load of fines after the fact. It is crucial that the regulations are applied in exactly the same way throughout the EU. The size of the internal market is our greatest strength for encouraging digitalisation, so greater predictability and clarity must be established about how you may and may not develop AI. The same interpretations should apply in all countries.
– Clear, simple and EU-wide rules would mean a lot for those start-ups and small businesses that lack the resources for complicated legal processes, but do have innovative ideas to contribute. It would just be good to have an EU-wide approach to AI technology. But it must be unbureaucratic and opportunity-oriented.
– Artificial intelligence is the rocket that will take us into the future, so we need to fuel it properly. Legislation must spur digital development and at the same time guarantee safety for all along the way.AIEU