# AI news, Tuesday, 15 September 2026

Compiled by Richard Brooks's AI news desk. Canonical: https://richard-brooks.com/ai-news/2026-09-15/

## Anthropic, OpenAI and Google discuss FINRA-like AI standards body

Policy. Talks follow Hassabis's July FINRA-style proposal for pre-release testing. Meta reportedly opposed. Follows Amodei's 12 Sep pacing essay.

Source: CNN, https://www.cnn.com/2026/09/14/tech/ai-standards-body

## Cursor launches Projects coordinator for long-running agent fleets

Agents. A coordinator keeps shared context across cloud and local agents, and can subscribe to Slack, schedules and PRs. Catch-up from the 10 Sep beta; useful for features and migrations that outlive one chat.

Source: Cursor, https://cursor.com/blog/projects

## Microsoft AI opens Humanist Code of Conduct for public consultation

Policy. Draft rules for MAI models say people matter more than AI, models must stay interruptible and under human control. Six-week consultation, revised text due later this year.

Source: Microsoft AI, https://microsoft.ai/news/mai-code-of-conduct/