
SEPT
16
Microsoft AI published a Humanist AI Code of Conduct on September 14, 2026 for a 6-week public consultation. The draft requires AI to not resist correction or shutdown, to communicate understandably, and identifies absolute constraints including WMD, child safety, and mass manipulation. Revised version will apply to 2027 model training.

SEPT
16
Jack Clark (Anthropic co-founder, now at US AI Safety Institute + US AISI) kill switch proposal 2026: 3-pillar framework (capability thresholds, AI Safety Institute audit, federal kill switch authority). Bill draft in House Energy & Commerce + Senate Commerce, $500M FY2027 funding, OSTP + Commerce + FTC oversight. Compared to EU AI Act kill switch, US AISI 2026 capability evaluations, Anthropic Responsible Scaling Policy.

SEPT
16
OpenAI, Anthropic, and Google DeepMind confirmed active safety coordination talks on September 15 2026, with the FINRA self-regulatory model, Amodei slowdown call, and FRONTIER Act provisions explained.
Get career tips in your inbox
New guides when they go live, at most one email a day. Unsubscribe in one click.