Trump, Beijing and Microsoft draw three lanes on AI control

Trump touted a “strong and smart” president, China called Amodei’s pitch “fearmongering” and Microsoft drafted a code.

Monday’s AI control fight split three ways: President Donald Trump said the only “guardrails” AI needs is a “STRONG AND SMART (High IQ!) PRESIDENT”; China’s Foreign Ministry branded Anthropic CEO Dario Amodei’s industry slowdown pitch “fearmongering”; and Microsoft AI put a draft Code of Conduct for MAI models out for public comment, including a hard line that models “will never resist human interruption, override, correction, or shutdown.”

The same 24 hours that put lab CEOs, Beijing spokespeople, and a hyperscaler code on the same beat also left the policy map uneven: White House messaging against new AI regulation, Chinese state security framing AI as strategic rivalry, and a U.S. platform promising “humanist” containment with feedback through an update that informs development starting in 2027.

Facts

Trump: “STRONG AND SMART” president as the guardrail

In a Monday Truth Social post, Trump wrote that the United States already has “tremendous CRIMINAL and REGULATORY power over these companies,” then: “The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” (Truth Social status 117269745153543631; archive trumpstruth.org/statuses/41712.)
Same-day posts, per CNN, cast a “SICK conspiracy” against AI and called safety concerns a “HOAX,” rejecting industry calls for more government oversight that have included Anthropic CEO Dario Amodei.
Secondary coverage (Axios, CNN) places the post against a White House line that has opposed new AI rules while some Republicans face voter skepticism on AI.

Beijing: MFA on Amodei; MSS on AI security and Party risk

Foreign Ministry spokesperson Guo Jiakun, at a regular Monday briefing, said: “Fearmongering, confrontation and malicious competition will only disrupt the process of global AI governance and serve no one’s interests,” responding to Amodei’s call for an industrywide slowdown (NBC News; MFA briefing attribution).
State-run Global Times, cited by NBC, blasted the slowdown pitch Sunday as “packed with containment provisions targeting China.”
State Security Minister Chen Yixin, in the state-run China Cyberspace magazine (Sunday), said China must “accelerate the establishment of a system for preventing and controlling AI security risks,” called for international cooperation, and framed AI as “a new track for strategic rivalry” (NBC). Separate secondary reporting (Reuters/Guardian circuit) says Chen also warned that AI could threaten Chinese Communist Party rule and named certain U.S. models as potential threats.

Microsoft: MAI Code of Conduct and Nadella’s “deliberate pacing”

Microsoft AI published “Humanist AI in practice: A public consultation on our Code of Conduct for MAI Models” dated September 14, 2026 (microsoft.ai), calling technology that does not serve humanity a failure and opening a first-draft Code for comment.
The Code states MAI models “will never resist human interruption, override, correction, or shutdown,” will “always recognize the primacy of human intent,” and will comply with pause, redirect, cancel, or shut-down requests under human-designed safety procedures (microsoft.ai/code-of-conduct/).
CNBC reports Microsoft CEO Satya Nadella wrote on X Sunday that “we welcome the research, focus, and deliberate pacing needed to get alignment right,” in the same window as Anthropic’s slowdown push; CNBC says an update will inform development starting in 2027.

Earlier coverage

Sep 7: Astra Critical / Nvidia–Hugging Face / EU AI Act RFIs.
Sep 8: Muse Spark / China SPC 法发〔2026〕10号 / Alien Mind metrics / Nvidia AI-factory MOUs.
Sep 10: Coxon exit + Hubinger >10% / OpenAI policy window + Newsom SB 813–AB 1405 / iPhone Duo from $1,999.

Meaning

Three power centers answered the same question — who sets the pace and who holds the kill switch — with incompatible answers: personal presidential control, state security plus MFA diplomacy, and a corporate code that treats human override as non-negotiable. The Amodei slowdown pitch is now a diplomatic and partisan flashpoint, not just a lab memo.

What’s next

Full China Cyberspace Chen text for the Party-rule and named-model clauses.
Whether Trump’s Truth Social line hardens into formal White House guidance vs. remains posting.
Microsoft consultation window and any 2027 MAI training-manual changes after comment.
Xi–Trump meeting timing (NBC notes AI approach as a likely agenda item for a September meeting).

— Lark Pemberton