Earlier this week I addressed the topic of AI regulation at the International Joint Conference on Neural Networks in Rome.
For me the key questions that I tried to address were:
How can we support innovation in Artificial Intelligence while protecting citizens and society from potential risks?
And how can we support this goal through regulation and policy at local, regional and global level?
At the outset we need to acknowledge that there are three broad approaches to the global regulatory landscape of AI. The EU, for example, has adopted a risk-based approach through the AI Act, which stands in contrast to the US free-market stance and China’s state-centric model.
Over the last four years, in my role within the Inter-Parliamentary Union (IPU), I have been directly involved in drafting the IPU Ethical Charter on Science and Technology. This charter aims to establish a regulatory framework of principles to guide legislation and decision-making, ensuring that all new scientific and technological innovations, including AI, benefit humanity, society, and the environment.
I have also been involved in delivering AI capacity-building workshops for parliamentarians and assisting with the drafting of the IPU resolution on the impact of artificial intelligence on democracy, human rights, and the rule of law.
Because these documents were adopted by 181 parliaments worldwide last autumn, they also serve as a valuable and accessible guide for the research and technology communities on the likely global approach governments and parliaments will take towards AI regulation.
In both instances the Charter and the Resolution identify a risk-based approach as the best mechanism for balancing innovation with regulation.
However, the single most important action the AI community can take right now is to help address the low level of AI literacy among parliamentarians and decision-makers. This will help foster an informed debate, leading to evidence-based decisions on AI policy and regulation.
But education alone is not enough; there is a clear need for early-stage, mutual consultation & engagement on the opportunities AI holds for society. This must be honest, two-way engagement, highlighting the positive opportunities in fields like medicine—where AI can support faster diagnosis and treatment, and even offer cures for disease—while also acknowledging the risks, such as the potential for harm when AI is combined with synthetic biology for malicious purposes, as highlighted by Mustafa Suleyman, co-founder of DeepMind, in his book The Coming Wave.
While bad news will always travel fast, we must also share the positive stories, and these shouldn’t just revolve around big tech, such as:
Small Business: In 2024, 13.5% of SMEs across the EU were using AI in their businesses. Take Lenehan’s hardware shop on Capel Street in Dublin, Ireland, a fifth-generation retail business that is using AI to provide advice to customers and help train new staff.
Health Service: The Irish health service, despite a historically mixed relationship with technology, is now embracing AI within its health app, being rolled out to citizens, and using AI to assist doctors in emergency departments to make better & more timely orthopaedic decisions from x-rays.
Highlighting the real, everyday opportunities that AI presents—particularly those which impact our daily lives—can help bring both the public and politicians on board, ensuring that Europe can lead, rather than follow, in this space.
But to achieve this goal we also need to see targeted investment in AI research and infrastructure to support innovation and address key challenges. For example, there are significant barriers to the use of AI within the EU, such as those highlighted by my Irish colleague Sean Kelly MEP regarding the need for long-term planning and critical investment in the EU electricity grid, as well as the problematic EU Commission definition of “high-speed broadband” as merely 100Mbps — something I fought to raise from an inadequate 30Mbps when I served on the Council of the EU.
So, where to from here?
This week, I suggested that the AI community take several steps, including:
Reaching out to parliaments in their respective countries to offer seminars explaining what AI is (and what it is not), its opportunities, and its potential risks.
Demonstrating how AI can be used to support parliaments and parliamentarians, by utilising the IPU guidelines for AI in parliaments.
Rather than trying to reinvent the wheel, organising a multidisciplinary forum with experts, policymakers, and decision-makers (parliamentarians and ministers) to review existing laws and regulations, identifying how they apply to AI and where gaps or loopholes remain.
Participating in Member State & European consultations and lobbying activities to shape the AI regulatory framework and ensure it caters for sector-specific needs.
Advocating for early-stage consultations between AI designers and a broad range of stakeholders—legal, psychological, social, and decision-makers—during the development of new AI systems, and promoting global collaboration at this stage to foster healthy competition.
Engaging in a more equal, global North-South dialogue on AI regulation and policy, to learn from diverse perspectives beyond Western institutions alone.
Share this post