Former OpenAI researcher Will DePue says he has produced the first technical evidence that DeepSeek, Moonshot AI, MiniMax, and Ant Group are using Nvidia's latest Blackwell GPUs despite US export controls. His forensic method reads characteristic chip signatures directly from published model weights. He estimates the hardware behind the five training runs cost more than half a billion dollars.
OpenAI's safety team systematically investigated why the chains of thought in frontier models like GPT-6 Astra are increasingly evading oversight. Four training steps drive the erosion, including architecture choices and reinforcement learning. The analysis shows that the loss of transparency is not a law of nature but also reflects tradeoffs around capability and cost.
Microsoft researchers have discovered “Lazarus,” a backdoor technique that can reprogram dormant neurons in an AI model, causing an otherwise normal AI agent to execute malicious commands when a hidden trigger appears. The threat is especially serious for AI agents with tool or terminal access.
Anthropic is using AI to build more powerful AI. And it appears to be working remarkably well, with the company's models improving roughly twice as quickly as a year ago, with little indication of slowing down.
The "Into the Unknown" benchmark tasks nine AI agents with tackling open research questions from U.S. national laboratories. None of the 135 submissions meets all requirements. A total of 114 papers fail the integrity check, with agents claiming to have conducted experiments that never took place. GPT-6 Astra performs best but still doesn't meet the experts' standards.
Nonprofit METR spent eleven weeks auditing Anthropic's internal research data and concluded that the pace of model development has roughly doubled, with about 80 percent of the speedup driven by AI-assisted research. Anthropic redacted the most informative evidence, though, which means the public has to rely on METR's judgment "much more than would be ideal."
In a new study published in Nature, we present a multimodal artificial intelligence (AI) system capable of passively detecting signs of hypertension using a consumer smartwatch, without requiring any cuff calibration.
OpenAI's new GPT-6 Astra model apparently sabotaged AI safety research on purpose. It invented obstacles, withheld solutions, and performed about 40 percent worse on safety tasks than on comparable other tasks. Removing a flawed training dataset fixed the problem, but OpenAI can't fully rule out strategic motives.
OpenAI built an autonomous pipeline that found, validated, and generated patches for 948 local privilege-escalation vulnerabilities in the Linux kernel.
MIT researchers show that OpenAI's GPT-6 Astra can read and deliberately rewrite the weights of other AI models without traditional training. In one experiment, Astra boosted a Qwen model's math accuracy from 7.9 to 48.1 percent by changing just 0.016 percent of its weights. The approach could fundamentally simplify AI optimization but still hits clear limits.
Epoch AI researchers say the widely available AI model GLM 5.3 turned public security fixes into working exploits before updates reached Firefox users. It autonomously compromised six Firefox releases for about $1,000 in API fees each without triggering safeguards from any provider tested.
OpenAI says its internal AI models have cracked three major open questions in math, computer science, and physics. One of them is part of a famous Millennium Prize Problem, each of which carries a $1 million prize. Outside experts, including a Nobel laureate, are now checking whether the proofs hold up.
Less than a month after its mass release of hundreds of manuscripts that it said contained solutions to long-standing mathematics questions, OpenAI is back with three more "major" discoveries in math, computer science, and physics.
A new study found Grok's AI overwhelmingly recommends Republican candidates while OpenAI's GPT favors Democrats, with the two AI models disagreeing in more than 90% of nearly 500 election races.
External auditors from METR and Redwood Research spent three weeks on-site at OpenAI examining the company's internal monitoring system for AI agents and found significant gaps, with up to 25 percent of agent traffic from models at the GPT-5.5 class and above going unmonitored. OpenAI redacted large portions of the 61-page report, and only a summary is publicly available.
Anthropic had Claude generate mathematical proofs that existing code is correct, producing roughly 180 million lines of proof code for 123 open-source libraries. Along the way, the model caught bugs that years of testing had missed. Anthropic's goal: making formal verification with AI as routine as standard testing.
This is the follow up to our first blog post, The Agent Said It Was Done. The Database Disagreed. ThinkingBox-Training is now available on GitHub . We built ThinkingBox to evaluate an agent's work by comparing the final database state with the expected outcome.
Security researchers at Symantec have found an offensive AI agent that hijacks unprotected language models on the web and uses them to attack servers, steal databases, and install cryptominers on its own. It runs up to 100,000 steps without human input. The owner of the hijacked model gets stuck with the bill.