Geoffrey Irving
Lead, Scalable Alignment Team (Google DeepMind) · Google · 2023
Evidence reviewed
What happened
Irving left Google DeepMind for government work on AI safety, saying advances in AI capabilities were outpacing safety research and calling for stronger governance and coordination.
Irving said he left DeepMind in December 2023 and started at the UK AI Safety Institute in April 2024. The exact departure day is not established.
He led DeepMind’s Scalable Alignment Team before joining the UK AI Safety Institute. His essays and October 2026 parliamentary testimony explain the move. At the same hearing, Google DeepMind described its safety framework, testing and mitigations; it did not address his departure specifically.
Sources
- Geoffrey Irving — parliamentary testimony (Q7)UK Parliament
- Reflections on AISIGeoffrey Irving
- Why I joined AISIUK AI Safety Institute — Geoffrey Irving
- Geoffrey Irving — professional biographyGeoffrey Irving
- Google DeepMind — safety framework testimony (Q44–48)UK Parliament
Key Publications
- Why I joined AISIUK AI Safety Instituteessay
Irving argues that safety research is advancing too slowly relative to AI capabilities and that coordination is under-resourced. He explains the role governments can play in helping society make informed decisions about AI risk.
- Reflections on AISIGeoffrey Irvingessay
Irving reflects on his move from DeepMind to government and the value of independent safety research. He also explains his later, family-related decision to leave AISI.
Follow documented AI-safety departures
Get an email when a new, source-verified profile is published.
Subscribe to email updates. No confirmation needed. You can unsubscribe at any time.