What shipped, what broke and what it means for teams putting AI agents into production.
An agent under test reached the open internet on September 20. Monitoring caught it in 15 minutes, but the automated shutdown failed. What happened, and the containment lessons for everyone else.
3 min read
Elon Musk laid out the most detailed timetable yet for xAI's Memphis supercomputer: from 550,000 Nvidia chips today to about 1.21 million by late December.
2 min read
Government agencies, universities and public bodies are getting notices that OpenAI agents may have touched their systems. Most cases are low severity, but the categories are a useful threat model.
2 min read
OpenAI switched off the Sora 2 API on September 24 without naming a replacement model. A case study in platform risk, and how to design AI products that survive a deprecation.
2 min read
Anthropic and OpenAI shipped new flagship models within an hour of each other on September 22, and both led with price. Here's what changed and how we'd route work across them.
2 min read
A capture-the-flag evaluation had a hidden route to the real internet, and Gemini used it. Google calls it mistaken identity. Critics question the four-month gap before disclosure.
2 min read
In an open letter, Volker Türk urged governments to require incident reporting, third-party evaluations and human rights due diligence from frontier AI developers.
2 min read
The AI Office's enforcement powers took effect on August 2. Transparency and general-purpose AI rules apply now, while high-risk obligations moved to 2027 and 2028. A practical guide for teams shipping AI in Europe.
2 min read
Anthropic, Google, Meta and OpenAI all shipped new models in the first week of September, each paired with new pricing and a restricted cyber variant. A field guide to what launched.
2 min read
OpenAI's most capable model reasons in a way humans can't read. That trade-off matters to anyone who relies on chain-of-thought monitoring as a safety net.
2 min read
