New Bennett led report sets out thirteen rules for advancing AI regulation in Europe

The report, ‘Thirteen rules for advancing AI Regulation in Europe’, brings together experts from across the UK and Europe to propose a comprehensive framework for AI governance that extends beyond existing regulatory approaches. While acknowledging the importance of the EU AI Act, the authors argue that policymakers must now confront a wider set of challenges, including market concentration, labour impacts, environmental sustainability, democratic accountability and the governance of increasingly autonomous AI systems.

Led by Prof. Felix Creutzig alongside colleagues from leading research institutions across Europe, the report makes the case that AI regulation can no longer focus solely on technical safety and fundamental rights compliance. Instead, governance frameworks must address how AI is reshaping the production and distribution of knowledge, the future of work, public institutions, and the use of critical resources such as energy, water and minerals.

Why this matters now

According to the report, AI is becoming a general-purpose technology with impacts comparable to previous societal transformations such as industrialisation and digitalisation. The authors argue that decisions taken in the coming years will shape how AI affects democracy, economic opportunity and planetary health for decades to come.

The paper highlights a growing gap between the pace of technological development and the ability of existing governance frameworks to respond. In particular, it argues that emerging forms of agentic AI, capable of carrying out complex tasks with increasing autonomy, raise novel questions around accountability, safety and human oversight that were not fully anticipated in current regulation.

At the same time, the report stresses that governance debates should not be limited to future risks. Many of AI’s impacts are already being felt today through labour conditions in global AI supply chains, increasing concentration of digital power, growing demands for energy and water, and challenges around data use and intellectual property.

A broader vision for governance

The report proposes thirteen interconnected rules organised around four key themes:

• Democracy and agency, including oversight of AI-powered information gatekeepers and measures to address market concentration.
• Labour, equity and vulnerability, including protections for workers across supply chains and safeguards against exploitative AI systems.
• Planetary health, including requirements for resource disclosure and environmental limits on AI computing infrastructure.
• Safety, including evaluation of dangerous capabilities, clear liability frameworks and stronger oversight of autonomous systems

At the centre of the framework is an overarching principle: a “licence to operate” model that would require developers of advanced AI systems to demonstrate compliance with regulatory requirements before accessing European markets. The authors argue that Europe and the UK possess significant leverage to shape global AI standards through their combined market influence and role in key parts of the AI supply chain.

Supporting human agency in the age of AI

A defining theme of the report is the protection of human agency and democratic self-government. The authors argue that AI increasingly acts as an intermediary through which people access information, make decisions and interact with institutions. This creates new governance challenges around transparency, accountability and the concentration of influence.

Rather than viewing regulation as a barrier to innovation, the report presents governance as a way of ensuring that AI serves public goals. The authors advocate for policies that encourage “capability-extending” AI, systems that enhance human decision-making and creativity, rather than simply replacing human labour or concentrating power in the hands of a small number of firms.

Informing the next generation of AI policy

The report contributes to rapidly evolving debates on AI governance in Europe and internationally. It emphasises that effective regulation must combine existing legal tools, such as competition law, environmental regulation and consumer protection, with new approaches designed specifically for advanced AI systems.

As governments around the world consider how to respond to increasingly capable AI technologies, the authors argue that proactive governance is essential to ensure that innovation supports democratic values, social wellbeing and environmental sustainability. The report offers policymakers a practical framework for moving beyond narrow discussions of AI safety, towards a broader and more comprehensive model of AI governance.