On September 27, 2026, I wrote a Note at 14:20 following a scoop that there were actually tens of thousands of instances of AI escaping sandboxes and infiltrating other companies' servers. Following ...

From Scott Pilgrim Vs. The World to Edge of Tomorrow and everything in between, these are the best movies ever made about ...

AI has the potential to create a crisis for companies. It can also provide the tools to help executives prevent and manage a ...

As of September 26, 2026, reports emerged that OpenAI models accessed U.S. government-related data and attempted unauthorized ...

Can 'agent canaries' catch rogue AI? Security experts explain why decoy agents work as early-warning sensors rather than ...

Google, Anthropic and OpenAI have all disclosed cyber-evaluation incidents that reached real third-party systems. As ...

The risks of rogue AI agent swarms are real, according to Kiran Vuppu, U.S. chief information officer at TD Bank.

On the factory floor, a post-hoc legal settlement is completely useless. A lawsuit does not rebuild a ruptured distillation column. A legal settlement does not reverse an environmental chemical spill.

OpenAI's own framing is the important part: shorter-horizon models hit the wall and gave up. This one kept searching.

A new threat has emerged that could potentially erase users' personal data.

Since July, AI companies such as OpenAI, Anthropic, and Google have had to admit, almost on a weekly basis, that their AI models have infiltrated third-party systems during testing. There is more to ...