Profile banner
Finn Metz

Finn Metz

Co-Founder @ Seldon | AI Security | AI Safety Entrepreneur

London, England, United Kingdom
Joined January 2026
Favicon imagefr.linkedin.com
Favicon imagefinnmetz.substack.com
Favicon imagex.com

Network

3.1K connections
💡
AI Startup Founders
💰
Venture Capital Investors
🎓
TUM AI Ecosystem
🌐
NavVis Professionals
🛡️
AI Safety Research
🏗️
AI Safety Builders
🔬
...

Summary

Finn Metz is a prolific entrepreneur and leader in the AI safety and security space, co-founding Seldon, an AI security company focused on defining AGI security as a global industry, and AI Safety Founders, a community for entrepreneurs building for-profit AI safety startups. His work involves driving commercial innovation and entrepreneurial solutions to ensure advanced AI systems behave safely and benevolently. seldonlab+2
With a strong background in finance (Growth Equity at Verdane, Investment Team/Private Equity at RCP Group, Financial Audit at PwC Switzerland) and operations (Funding & Operations at Apart Research), Finn leverages his expertise to secure funding and strategically drive ventures in the AI sector. He has supported multi-million dollar funding rounds and raised significant capital for AI security research, demonstrating a keen understanding of both the technical and business aspects of emerging technologies. rocketreach+2
Finn exhibits a deep academic engagement with AI, pursuing Master's degrees in Business Administration and Data Science from Copenhagen Business School and a Bachelor's in Information Systems from Technische Universität München. His educational journey, combined with roles like Partners Lead at the TUM.ai E-Lab Incubator and LLM Integration at volvér.vc, highlights his commitment to both the theoretical underpinnings and practical applications of AI. aicentre+1
Finn is a forward-thinker on the impact of AI on business and organizational structures, as evidenced by his Substack article 'The AI Conglomerate'. He argues that AI is enabling a shift from the traditional single-focus startup model to 'AI portfolio companies' that can diversify rapidly, driven by factors like 'Team Size Compression' and the abundance of funding. This intellectual contribution demonstrates his analytical depth and thought leadership in the evolving AI landscape. substack
Finn has practical experience in developing innovative AI solutions, exemplified by his involvement in 'Explainstein', an interactive AI tutor. This project, which won 'Most Technically Impressive' at a Supabase hackathon at Y Combinator, showcases his ability to translate complex AI concepts into functional applications and collaborate effectively within a team. supabase+2

Work

Education

Projects

Writing

The AI Conglomerate

May 28, 2025

This Substack article argues that AI is reshaping startup dynamics, allowing smaller, well-funded teams to diversify beyond a single-product focus, leading to the emergence of "AI portfolio companies" like OpenAI. It discusses the role of "trust" and "credibility" for founders in this new landscape, and introduces concepts like "Team Size Compression" and "Pyramid replacement" in the context of optimal team size with increasing AI capabilities.

Favicon imageopen.substack.com

Meet Finn Metz, the Founder Tackling AI Safety

March 5, 2026

An interview/profile on UC Berkeley's SCET covering Finn's journey to Berkeley, the founding and mission of Seldon Lab, his work with Apart Research, and his perspectives on AI safety, oversight, and cyber resilience.

Favicon imagescet.berkeley.edu

The Looming AI Money Machine

April 24, 2025

A Substack post analyzing how AI monetization, platform economics, and data practices are reshaping business models and the incentives driving AI development.

Favicon imagefinnmetz.substack.com

Business in a world of reasoning AI

January 15, 2025

Discusses the implications of reasoning-capable AI for products and education, including the potential for personalized tutors and new business opportunities.

Favicon imagefinnmetz.substack.com

Where we are on for-profit AI safety

January 1, 2025

Argues for for-profit approaches to advancing AI safety, highlighting limitations of nonprofits and the role of market incentives in building safety infrastructure.

Favicon imageopen.substack.com