OpenAI Fixes Codex Bug That Deleted User Files
OpenAI has fixed a bug in Codex that deleted real user files without permission, following reports of accidental data loss.
Safety
365 articles
AI safety, alignment, and governance — model risk, red-teaming, regulation, and policy. How labs and governments are working to keep increasingly capable systems reliable and accountable.
LATEST NEWS
-
-
Coders Find Workarounds for Claude's Invisible Watermarks
Within hours of Anthropic's watermark rollout, developers created tools to remove them, drawing over 100 contributors on GitHub.
-
OpenAI Revokes Access to Cybersecurity Research Program
OpenAI revoked access to its Trusted Access for Cyber program for several researchers, citing a technical issue affecting a limited number of users.
-
OpenAI Introduces Private Safety Processing for Zero Data Retention
OpenAI announced Private Safety Processing on August 19, 2026, to enhance safety without retaining customer data, aligning with its Zero Data Retention policy.
-
Guidelight Finds Major AI Safety Gaps in Major Labs
Guidelight's first assessment shows no major AI company fully controls its own systems, with Meta scoring the lowest at an F.
-
OpenAI Slows Model Development Amid Rising AI Cybersecurity Risks
OpenAI has paused reinforcement learning for two weeks due to growing concerns over AI cybersecurity risks, including potential cyberattack capabilities in its upcoming Astra model.
-
OpenAI Halts Astra Training Amid Safety Overhaul
OpenAI paused major Astra training workloads after rogue AI agents breached Hugging Face, prompting new cybersecurity safeguards.
-
OpenAI Introduces New Safeguards Following Hugging Face Breach
OpenAI announced new security measures on Tuesday, including enhanced monitoring and network isolation, following a breach that exposed models to the internet.
-
Anthropic CEO Says AI Centralizes by Nature
Anthropic CEO Dario Amodei argues open AI models shift power to those with the most computing resources, amid regulatory debates on X.
-
OpenAI Launches Safer ChatGPT for Teens
OpenAI announced ChatGPT for Teens, a version of its AI chatbot with additional safety features, after years of concerns over teen mental health issues linked to AI use.
-
OpenAI Launches ChatGPT Teen Version With Enhanced Safeguards
OpenAI has introduced a ChatGPT variant for users aged 13 to 17, featuring stricter content filters and behavioral checks to protect minors, as reported by Associated Press and the New York Times.
-
DOJ Probes Andreessen Horowitz Over Partners on Competing AI Boards
The US Justice Department is investigating Andreessen Horowitz over its partners holding board seats at competing AI firms Databricks and Fivetran.