0:00
/
0:00

Eoghan O’Neill, Senior Policy Officer at the European Commission AI Office — Unpacking the AI Act & Europe's Global AI Play

In this episode, Andreas Munk Holm is joined by Eoghan O’Neill, Senior Policy Officer at the European Commission’s AI Office, to break down the EU AI Act, Europe’s strategy to lead the global AI wave with trust, safety, and world-class infrastructure.

They dive into why the EU’s approach to AI is not just regulatory red tape but a proactive framework to ensure innovation and adoption flourish across sectors—from startups to supercomputers. Eoghan unpacks how startups can navigate the Act, why Europe’s regulatory clarity is an advantage, and how investors should be thinking about this new paradigm.

Here’s what’s covered:

  • 02:41 Eoghan’s Unorthodox Journey: From Gravy Systems to AI Policy

  • 04:32 The Mission & Structure of the AI Office

  • 05:52 Understanding the AI Act: A Product Safety Framework

  • 09:40 How the AI Act Was Created: An Open, 1,000+ Stakeholder Process

  • 17:42 What Counts as High-Risk AI (And What Doesn’t)

  • 21:23 Learning from GDPR & Ensuring Innovation Isn’t Crushed

  • 26:10 Transparency, Trust & The Limits of Regulation

  • 30:15 What VCs Need to Know: Obligations, Timelines & Opportunities

  • 34:42 Europe's Global AI Position: Infra, Engineers, Strategy

  • 43:33 Global Dynamics: Commoditization, Gulf States & the Future of AGI

  • 48:46 What’s Coming: Apply AI Strategy in September

📺 Watch the episode here or stream it on Spotify or Apple Podcasts—now with chapters for easy navigation 🎧


✍️ Show Notes

The AI Act: Europe’s Bold Bet on Safety, Trust & Adoption

Far from stifling innovation, the EU AI Act aims to unlock it by offering clarity and trust in AI systems. Think: escalators, aviation, medical devices—AI is joining the club of tech governed by mature, proven product safety regulation.


Who’s Regulated, and When?

There are two main “populations” the Act applies to:

  • Generative AI model providers (e.g., OpenAI, Mistral) — compliance begins August 2025

  • AI systems builders (i.e., most startups using AI) — must understand risk tiers (prohibited, high-risk, low-risk) and align accordingly

A public consultation and clear guidelines are underway to make this easy to navigate, with final guidance expected February 2026.


Defusing Regulatory Capture

The Commission isn’t dictating rules from an ivory tower. With 1,000+ stakeholders involved—ranging from Anthropic and OpenAI to NGOs and SMEs—the Code of Practice is being shaped by the ecosystem. It’s chaired by figures like Yoshua Bengio, with a clear focus on avoiding big tech dominance.


Transparency Is the New Trust Layer

Just like a box of cornflakes, AI systems in Europe will soon come with ingredients and “nutritional” labeling. This transparency layer can drive market adoption and serve as a moat for startups operating in regulated markets.


For VCs: Europe’s a Market Maker, Not a Market Killer

  • The AI Act creates a common framework across 27 countries

  • Most startups fall outside high-risk, facing minimal obligations

  • It creates opportunities in verticals like healthcare, law, and education

  • Legal clarity helps de-risk investment


Europe’s Strategic Advantage: Infra + Talent + Policy

  • AI Engineers: 30% more per capita than the US, 3x China

  • Supercomputers: 15 of the global top 50 are in the EU

  • Funding: €20B AI Gigafactories + €1B+ AI Factories investments

  • Free compute for approved startups launching H2 2025


Apply AI Strategy: Removing Friction to Scale AI in Verticals

Launching in September 2025, this will outline how to scale AI across verticals—based on real-world use cases and policy actions to overcome blockers like data access, compliance bottlenecks, and institutional inertia.


Why VCs Must Engage Now

Too many are passive or misinformed—often echoing US narratives that misunderstand the role and scope of EU regulations. The Commission wants to hear from the venture ecosystem. Think of it as a two-way street.


EU AI Deep Dive

The EU is laying serious groundwork to own the AI era—not just in legislation, but also infrastructure, strategy, and adoption. Here’s why we think Europe’s positioning is underappreciated:

  • Legal Clarity: 1 regulation, 27 markets

  • Infra Access: Public compute resources, free for startups

  • Talent Base: AI engineers and researchers distributed across leading institutions

  • Strategic Foresight: Balancing innovation and safety from Day 1

Open questions remain on defensibility, global regulatory convergence, and how fast agentic AI will scale—but the foundations are being laid now.

Connect with Europe’s leading LPs, Family Offices & Fund of Funds this September at Château de Beloeil.

🤗 Join the EUVC Community

Looking for niche, high-quality experiences that prioritize depth over breadth? Consider joining our community focused on delivering content tailored to the experienced VC. Here’s what you can look forward to as a member:

  • Exclusive Access & Discounts: Priority access to masterclasses with leading GPs & LPs, available on a first-come, first-served basis.

  • On-Demand Content: A platform with sessions you can access anytime, anywhere, complete with presentations, templates, and other resources.

  • Interactive AMAs: Engage directly with top GPs and LPs in exclusive small group sessions — entirely free for community members.

Join the EUVC community

🗓️ The EU CVC Summit 2025 | 📆 26 - 28 August 2025 | 🇩🇰 Copenhagen, Denmark