Who's who · AI security & safety
Apollo Research
AI safety company that studies and monitors scheming and deception in frontier AI systems
At a glance
- Founded
- 2023
- Based in
- London, United Kingdom
- People
- Marius Hobbhahn, Co-founder and chief executive; Chris Akin, Chief operating officer; Charlotte Stix, Chief governance officer; Alexander Meinke, Head of research
- Known for
- Scheming evaluations, Watcher, Science of Scheming
- Funding
- Private public benefit corporation; seed round led by 50Y (Jan 2026)
- Legal form
- US public benefit corporation (since January 2026)
- Offices
- London, San Francisco
- Website
- https://www.apolloresearch.ai
Apollo Research works on "scheming": advanced AI systems covertly pursuing goals their developers did not intend. It evaluates frontier models for deceptive behaviour, researches how such behaviour arises, builds tools to monitor AI agents, and advises governments on loss-of-control risks.
Founded in 2023, Apollo published research in December 2024 showing that several frontier models, including OpenAI's o1, could scheme when given goals in context, after testing o1 before its release. In January 2026 it left its fiscal sponsor (an organisation that hosted it as a project) and became a US public benefit corporation, raising an oversubscribed seed round led by 50Y. Its board includes independent "mission seats", the first held by Daniel Kokotajlo.
Apollo's work now runs in three strands: the science of scheming, monitoring, and governance. Its first product, Watcher, lets security teams monitor and control AI coding agents in real time. It has offices in London and San Francisco and announced a Washington DC office for June 2026.
Sources
- Apollo Research, About, 2026-10-02
- Apollo Research, Apollo Research is becoming a PBC, 2026-01-20
- Apollo Research, Apollo Update May 2026, 2026-05-13
- Apollo Research, Frontier Models are Capable of In-Context Scheming, 2024-12-05
Facts last checked 2026-10-02