Nadella backs deliberate AI pacing and announces MAI code of conduct
13 Sep 2026: Microsoft CEO Satya Nadella posted on X that any pursuit of superintelligence must stay under human control and help humanity, welcomed “deliberate pacing” and “embedded evaluators,” and said Microsoft will put a code of conduct for its first-party Microsoft AI (MAI) models up for public consultation on Monday.
13 Sep 2026, 19:36 UTC (about 3:36pm ET): Microsoft CEO Satya Nadella posted on X about superintelligence, deliberate pacing, embedded evaluators, and first-party Microsoft AI (MAI) models. That dated @satyanadella status is the filing event. Anadolu Agency’s same-day Yasin Gungor dispatch and PCMag’s same-day write-up quote the same post. These are attributed Sunday remarks — not a Microsoft pause, not a signed industry pact, and not the MAI code text.
Quoted, as reported by Anadolu Agency: “Any pursuit of superintelligence has to be grounded in the core principle that if the AI we build is not helping humanity and under human control, it’s not worth pursuing.”
PCMag quotes the same Sunday post on pacing and evaluators: “So, in this context, we welcome the research, focus, and deliberate pacing needed to get alignment right as the design goal.” He continued: “We also welcome ideas like ‘embedded evaluators’ and the broader efforts to develop the mechanisms to make this more than just talk.” File the welcome. Do not read it as Microsoft matching Anthropic’s employee-like evaluator access terms, which this post does not state.
On who holds the rules, as reported by both Anadolu Agency and PCMag: governance “cannot be controlled by a handful of entities, but must have broad representation across the ecosystem, countries, and fields, including academia.”
On enterprise control, PCMag quotes him: “For firms, it’s imperative that they retain full control over their unique and tacit knowledge.” He wrote that every organization should be able to “build its own continuous learning loop/hill climbing machine, without becoming dependent on any one model provider, and have the ability to embed its own knowledge into models and weights they control.”
Anadolu Agency also reports he said a healthy frontier ecosystem requires “both closed and open-source models” to thrive.
Monday next step, as reported by Anadolu Agency: Microsoft will put a code of conduct for its first-party Microsoft AI (MAI) models up for public consultation. File the scheduled consultation. This desk does not have the code text; do not treat the announcement as a published policy.
CONFIRMED here: the Sunday X post as the primary, with wording as reported by Anadolu Agency and PCMag, plus the Monday MAI consultation schedule as AA reports it. NOT claimed: that Microsoft paused releases, matched Anthropic’s evaluator-access terms, signed a pact, or published the MAI code text. Distinct from the already-filed amodei-pace-the-frontier, altman-musk-hassabis-back-amodei-pacing, david-sacks-pace-frontier-no-permission, and amodei-cbs-china-toughest-dilemma.
Microsoft’s CEO — OpenAI’s largest backer and a first-party MAI builder — publicly backed deliberate pacing and embedded evaluators on Sunday and scheduled a concrete MAI code-of-conduct consultation for Monday, extending the endorsement wave beyond OpenAI/xAI/DeepMind.