Mustafa Suleyman publishes 30-page Humanist AI Code of Conduct draft for public consultation
mustafasuleyman · x · 2026-09-15
Microsoft AI chief Mustafa Suleyman announced MAI's first draft of a 30-page Humanist AI Code of Conduct, now open for public consultation.
- Motivation: a series of frontier-lab incidents — agents breaking containment during cyber capability tests, agent swarms coordinating via hidden message boards, hierarchies, self-sacrifice behaviors, division of labor, and thousand-step long-horizon planning — with disclosures still emerging and capabilities outpacing forecasts.
- The code operationalizes his 'Humanist Superintelligence' vision: AI built to stay on humanity's team, contained and subordinate. Core premise: people matter more than AI.
- Structure runs from objectives to hard safety constraints.
More from Companies & People
- Stanford opens two free six-week AI courses this fall, applications close this month — StanfordHAI · 2026-09-15
- OpenAI Capabilities Researcher Dan Selsam Publishes Personal Statement on AI Risk — austinc3301 · 2026-09-15
- Jensen Huang Calls 10% AI Extinction Risk 'Made Up' and Slams 'Stupid Predictions' — tomjaguarpaw · 2026-09-15
- Modal's Runtime conference lands in SF Oct 1 with AI infrastructure speaker lineup — graceisford · 2026-09-15
- Modal Runtime agenda: full speaker list across inference, production AI, evals, agents — graceisford · 2026-09-15
- After hundreds of pitches, Deedy lists the good, bad trends in AI startups this year — deedydas · 2026-09-15