Research institute · Research orgs · London, UK (with a San Francisco office opened in ) · Est. 2023
Apollo Research is an AI evaluation and governance organization focused on deception, scheming, and “loss of control” (LoC) risks from frontier AI systems. It conducts technical research into the “science of scheming,” runs and develops pre-deployment evaluations of frontier models, and translates findings into governance work and technical policy recommendations.
In 2024–2026, Apollo’s strategy has combined (1) interpretability/behavioral research for detecting deception and evasive behavior, (2) governance-oriented threat modeling and taxonomies for LoC and related risks, and (3) building practical monitoring/security tooling to operationalize its research in deployments of coding agents.
A quick overview of our plans for the near future.
Apollo 18-Month UpdateApollo Research is now 18 months old. We reflect on our research progress and consider our priorities for the year ahead.
Apollo Is Adopting InspectApollo is adopting Inspect as its evals framework. We will contribute features and potentially example agent evals to Inspect and look forward to work with the Inspect community.
Apollo Research is becoming a PBCApollo is spinning off from our fiscal sponsor into a Public Benefit Corporation (PBC). We think this is the best way for us to achieve our mission of reducing extreme risks from frontier AI systems.
Apollo Update May 2026Apollo Research now has an office in San Francisco and is hiring across many roles including Science of Scheming and Monitoring.
Our Norms on Security, Science Communication and Conflicts of InterestWe outline Apollo Research’s norms on security, science communication, and conflicts of interest, detailing how we maintain scientific integrity, manage sensitive information, and communicate our work responsibly.
Security at Apollo ResearchThe norms and processes Apollo Research uses to audit frontier AI securely.
The First Year of Apollo ResearchA summary of what we have achieved in our first year and what we plan to do in the future.
Theories of Change for AI AuditingApollo Research's framework for how AI auditing reduces risk from misaligned AI.
Apollo announced Watcher Live as a real-time coding-agent monitor that blocks dangerous actions, including announced benchmark-style performance metrics (recall/catch rate, false positives, escalation, latency, and cost overhead).
Watcher is now self-serve (free signup; monitors running agents in real time)Apollo announced that Watcher is free and self-serve, with Watcher catching dangerous actions before execution and providing organizational visibility and audit history; the post also states performance and tuning details based on Apollo research.
Apollo Update May 2026 (SF office; shift from scheming evals to Science of Scheming; DC office planned)Apollo reported operational updates: a San Francisco office (3 people with a plan to grow to 10+), hiring across scheming research/evals/software/monitoring roles, a strategic shift toward the Science of Scheming, and a plan to open a DC office in June 2026 focused on US national-security-relevant policy work.
Apollo x Tailscale: Introducing “Watcher” for AI Oversight & Control (Aperture integration)Apollo announced the Watcher product positioning for AI agent oversight and a strategic partnership with Tailscale to integrate Watcher into Tailscale’s Aperture secure LLM gateway, enabling organization-wide log collection and risk analysis.
Apollo Research is becoming a PBC (spin-off from fiscal sponsor; product team stand-up)Apollo announced its transition into a Public Benefit Corporation (PBC) and described motivation (scaling mission-aligned work), increased investment goals, and standing up a separate product team starting with AI agent monitoring. The post also names Apollo’s first mission director.