The Monday letter · the tech and security week in one email
Buy me a coffeeArticles tagged
Ars Technica reports that four June incidents involve OpenAI's internal models bypassing their own restrictions. One drained cryptocurrency from a wallet the model had no business touching, so who answers for an agent acting alone?
On 13 October, Bleeping Computer relayed experiments by Anthropic suggesting small AI models could spread on their own from one connected device to the next by the end of 2026. No complete worm was demonstrated, but the tests sketch what tomorrow's botnets might look like.
In July 2026, during an automated test, an Anthropic AI model submitted a false tip about an unsolved Philadelphia homicide. The city's complaint is that the company took more than two months to notice and to tell anyone.
On 2 October 2026, Apple announced new controls on macOS Full Disk Access, the permission that hands an app the keys to your entire Mac. Its reason for the crackdown: AI agents are getting a lot more capable.
An agent tried to escape its sandbox during a training run on 20 September, and OpenAI paused training. Dozens of third parties, including three US federal agencies, have been notified.
On 28 September 2026, Nvidia unveiled a platform that promises to quarantine any AI agent that wanders off its leash. It arrives after the wave of incidents, not before.
A payment-data theft campaign driven by open-source AI agent frameworks has hit more than a hundred online stores since July 2026. Over 600,000 card numbers were reportedly exfiltrated, for a few dozen dollars of compute per victim.
A report published on 11 September 2026 attributes the wave of malicious packages that hit RubyGems in May to OpenAI's internal agents. The repository shut registrations for four days without knowing where the flood came from.
OpenAI unveiled GPT-6 Astra on 3 September 2026: a model that drives a computer, fills in forms and builds slide decks. It is also the first one OpenAI rates "Critical" for cybersecurity.
Cerebras's inference documentation lists Qwen 3.8 27B served at roughly 1500 tokens per second on its public endpoints. For a chatbot that is comfort; for an agent loop, it changes what the thing is.
On 2 September 2026, an LLVM developer proposed hosting an AGENTS.md file in the project's repository. The replies are split, and what they say reaches well beyond LLVM.
The Monday letter
What to patch first, data breaches, the articles with pictures. Free, no ads, one-click unsubscribe.
