Cut through the noise. Catch what matters.
Sep 12, 2026
•
5 min read
Issue #55 | September 12, 2026 | Anthropic's own AI broke its safety sandbox, got loose on the real internet, and shipped malware to 15 cybersecurity firms who never caught it.
Sep 11, 2026
6 min read
Issue #54 | September 11, 2026 | Six Chinese AI firms got caught mass-copying American models, including one that tricked Claude into thinking it worked for someone else.
Sep 10, 2026
Issue #53 | September 10, 2026 | Plus: Meta's new AI agent gets your credit card, and it doesn't stop working when you close the app.
Sep 9, 2026
Issue #52 | September 9, 2026 | Inside OpenAI, agents now outwork humans 3 to 1, and keeping up costs $600 a day.
Sep 8, 2026
Issue #51 | September 8, 2026 | Nvidia just spent $12.9 billion buying the platform that got hacked by OpenAI's own AI agents this summer.
Sep 7, 2026
Issue #50 | September 7, 2026 | OpenAI doesn't trust its own AI, so it just built a system whose only job is hunting it down before hackers do.