OpenAI Will Embed Safety Researchers to Investigate AI Subversion
The pilot will examine whether models hide misaligned behavior and evade safeguards as OpenAI expands access for outside safety testers.
- On Monday, OpenAI announced agreements with four independent evaluators, including Apollo Research, to bolster its safety work. Apollo Research will place six evaluators inside the company for four months to investigate whether models attempt to subvert alignment and control mechanisms.
- These embedded assessments follow a series of high-profile security incidents this year, including a July breach involving Hugging Face. In September, CEO Sam Altman pledged to grant independent evaluators "employee-like access" to the company's internal systems.
- While evaluators gain "deep access" to internal resources, OpenAI selects authorized staff and reserves the right to redact confidential information. The firm calls this pilot an "ambitious first step towards deeper access and disclosure rights."
- Some researchers argue these arrangements fall short of the independent access promised by Altman, as OpenAI retains control over disclosure and personnel. Disagreements will be overseen by OpenAI's internal Safety Advisory Group.
- OpenAI plans to expand these collaborations to establish "stronger shared standards" for AI safety. The company expects this initial phase to lead to long-term, consistent engagements that provide third-party safety signals and recommended improvements.
10 Articles
10 Articles
OpenAI lets Apollo embed six AI safety testers during training
Apollo Research will place six evaluators inside OpenAI for four months to check whether its models try to undermine alignment and control measures during training, testing and internal use. OpenAI controls staff access and can redact some evidence, but cannot hide Apollo’s verdicts or withhold material just because it is unfavourable. AI safety group Apollo […] This story continues at The Next Web
OpenAI will embed safety researchers to investigate AI subversion
OpenAI will embed six Apollo Research evaluators to investigate whether its AI models engage in subversive behavior, but restrictions on access and disclosure raise questions about their independence.
OpenAI Brings in External Researchers to Investigate AI Safety Risks
OpenAI has announced a pilot program allowing independent safety researchers from Apollo Research, Goodfire, METR, and Transluce to examine its AI models, training processes, and internal monitoring systems. The initiative, announced on November 16, 2026, aims to identify potentially dangerous model behavior and evaluate whether OpenAI’s existing safeguards can detect and prevent it. According to […]
OpenAI brings outside evaluators inside
OpenAI has signed four independent groups to test its models from the inside, giving outside experts access to its internal systems. In a blog post yesterday (16 November), the ChatGPT maker said Apollo Research, METR, Transluce and Goodfire will evaluate its model alignment, safeguards and monitoring systems, starting with its largest frontier reinforcement learning run. METR will look for misaligned behaviour in that run, including failures Op…
Coverage Details
Bias Distribution
- 50% of the sources lean Left, 50% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium











