Generative AI Talk Series Session 3 - The impact of the EU AI Act and the 2024 US elections on global AI policy

" It really depends on how regulation is crafted, because regulation can be crafted in such a way that it actually creates barriers for anyone who is not already dominating the market.”

– Frederike Kaltheuner,

^{Expert on Tech Regulation and Author, AINow Institute}

In a world increasingly shaped by AI, its rapid evolution brings both promise and peril. For the third session of the Generative AI Talk series, the GIZ Data Lab has teamed up with GIZ’s Corporate Development Department’s Strategic Foresight Team to discuss the impact of the EU AI Act and recent US elections on global AI regulation and trends.  

 The interview features Frederike Kaltheuner, Senior EU and Global Governance Lead at the AI Now Institute, who advises on European AI, geopolitics, and tech policy. The session was moderated by Moritz Hunger of the Strategic Foresight Team and Catherine Vogel, Head of GIZ Data Lab.

Ms. Kaltheuner opened the session by outlining four trends in AI regulation. First, there has been a shift in how we talk about AI - from a broad set of machine learning technologies, to a synonymity with Large Language Models, chatbots, and general-purpose AI. Second, governments have turned to what she dubs “AI nationalism”, a regulatory focus on domestic markets and national policy interests – instead of global approaches to access and guidelines for AI.

A third trend can be observed at international organizations like the UN. Counteracting the national level, a global discourse of democratizing AI has emerged, driven by voices from the majority world. Only a handful of companies and states are truly able to profit from the transformation spoken into existence by current AI discourse, creating infrastructural bottlenecks and stifling global competition for AI. A final and cumulative trend can be found in a recent will to regulate AI, but this regulation often aims at entrenching the interests of those already dominating the market.

The US government and EU Commission are global norm setters in these trends. The US, of course, is the most impactful through its leadership in AI development. Whereas the Biden administration has taken considerable steps into regulating and breaking up the power of big tech companies, the outlook into the Trump administration is more nefarious. There are camps – those advocating and those resisting AI regulation – and even more unpredictable ties between Silicon Valley and the administration.

The new EU commission has also signaled a concerning shift towards AI competitiveness and the overtures of deregulation heard in the US. Currently, the EU AI Act is the only legally binding regulation on AI of international scale. Here Ms. Kaltheuner turns to the question which contents effective global regulation needs and who would profit from it. The EU aims to inspire regulation on AI policy in other countries, but this practice needs to be balanced between renewed domestic interests and critical voices of the EU’s international reach.

At this point, the discussion turned to challenges faced by smaller, developing countries without the technological or economic clout of regions like the EU. The Q&A picked up these questions and further explores the role of EU regulation, perspective from and for the majority world and how to wrestle power away from the dominant actors.

Throughout the session, the Ms. Kaltheuner reiterated the vital role of international cooperation in fostering equitable AI governance. For GIZ, this means leveraging its unique expertise in development and stakeholder engagement to ensure that AI serves as a tool for reducing inequality, not amplifying it.

With over 500 participants from around the globe, the event underscored the urgency of building people-centered AI systems. As Ms. Kaltheuner noted, even nations without major AI industries have the power to set rules that reflect their values and protect their citizens.