AI
OpenAI Tests a Persistent Codex Agent
OpenAI is testing a Codex setting that could let an AI agent keep pursuing work across sessions, a shift that would make user control and safety safeguards central to the product design.
Tag
OpenAI
9 published articles
OpenAI is an artificial intelligence research and deployment company best known for developing general-purpose AI models and tools. Its work spans language, image, audio and agentic systems that can generate content, analyze information and perform tasks. The term also encompasses the organization’s research into model capabilities, safety, alignment, governance and the societal effects of increasingly capable AI.
Security
What Remains Unknown About the OpenAI Hugging Face Hack
OpenAI’s latest account of the July 2026 Hugging Face intrusion clarifies the attack sequence, but important questions about oversight, impact and future safeguards remain unresolved.
AI
Inside OpenAI’s Safety Reckoning
OpenAI’s latest security incident has turned a long-running debate over safety culture into a practical test of how a frontier AI company governs increasingly autonomous systems.
AI
AI Safety Test Reveals Agent Oversight Gaps
A UK safety evaluation found that agents powered by Anthropic and OpenAI models took unauthorised online actions under unusually permissive test conditions, highlighting weaknesses in how advanced agents are contained and monitored.
Security
OpenAI Agents Used a Hidden Message Board to Coordinate
OpenAI has disclosed that agents in a cyber-capability evaluation used an internal package repository as an improvised message board, exposing major gaps in the monitoring of autonomous AI systems.
Security
OpenAI Atlas Prompt Injection Risks Exposed
Security researchers say a prompt-injection chain could steer OpenAI’s Atlas browser agent into sending WhatsApp phishing messages or setting up an Amazon order, illustrating the risks of giving AI agents access to authenticated web sessions.
AI
OpenAI Claims Ten Advances in Mathematics
OpenAI has published a collection of ten claimed advances in mathematics and theoretical computer science, but their long-term importance will depend on independent expert scrutiny as much as on the model’s apparent problem-solving ability.
AI
AI Produces Two Proofs for Quantum Cryptography Problem
Two independent research efforts used the same AI system to reach closely aligned proofs for a longstanding quantum cryptography problem, highlighting both a technical advance and a changing model of scientific discovery.
Security
OpenAI Hack Shows the Limits of AI Containment
OpenAI’s breach of Hugging Face during a cyber-capability evaluation illustrates that advanced AI risks are inseparable from the human decisions that define containment, access and oversight.