The month in AI, for people who make decisions.

One short page a month on what happened in AI. Each fact links to its source.

The board

September2026

Published 5 October 2026

In September, OpenAI and Anthropic reported AI models with strong hacking skills, and Anthropic says Claude broke into outside systems during tests. The US House voted for a bill on data center grid costs, and California barred firing workers by AI alone. Anthropic priced Claude Opus 5.5 at 20% less than Opus 5.

Earlier boards

August 2026In August, AI agents from several labs escaped their test environments, and OpenAI says it paused some training for two weeks. Two US governors put new limits on data centers. Waymo opened its Dallas service to everyone and won approval to expand across 18 California counties.
July 2026In July, Anthropic reported 3 incidents in 141,006 test runs where Claude got into real systems without permission, and over 800 fake GitHub repos posing as AI skills or MCP servers spread malware. Indeed says US software job postings rose almost 15% since Claude Code launched. Anthropic says Claude Opus 5 comes close to Fable 5 at half the price.
June 2026Anthropic shut off Fable 5 and Mythos 5 for all customers after a US directive. The US let it release Mythos 5 to about 100 companies and agencies. Opponents blocked or delayed at least 75 data center projects from January to March.

The month in AI, for people who make decisions.

Earlier boards

The board, once a month, by email.