AI News, 3 October 2026
The 1 most significant AI story of the day, selected and summarised automatically by AI. Each links to the original reporting.
- Policy
OpenAI Safety Staffer Quits and Goes Public
auto_awesomeAI SummaryDavid Robinson, who authored safety evaluations for every major OpenAI model release, has resigned and published a critical editorial in The Atlantic. His departure adds to a growing pattern of safety-focused employees leaving OpenAI and raising public concerns. The move intensifies scrutiny on whether OpenAI is adequately prioritising safety as it accelerates model development.
Why it matters: Each high-profile safety departure chips away at public and regulatory confidence in OpenAI's internal governance at a critical moment when governments worldwide are actively shaping AI policy. Robinson's insider credibility is significant — he was not a peripheral voice but the person responsible for formal safety documentation on flagship models. If safety evaluators themselves believe the process is inadequate, it raises hard questions about what those reports actually captured. Regulators, enterprise customers, and partners who rely on OpenAI's safety disclosures should be paying close attention to the pattern of who is leaving and why.