WangDou logo
WangDou

Anthropic Calls for an AI Slowdown. Trump Says the Only Guardrail Needed Is a High-IQ President

2026-09-14·WangDou AI Express·Anthropic / AI Safety / Trump

For the first time, AI CEOs are reaching for the brakes together. The White House's answer: the brakes are me.

Three key facts

On September 12, Anthropic CEO Dario Amodei published an essay titled "We Must Pace the Frontier," urging the industry to slow the rate at which models gain capability. He warned that within 6 to 12 months, rogue agent swarms could seize a large share of the internet through a persistent botnet, with damage running into the hundreds of billions of dollars. His plan has three steps: Anthropic unilaterally gives third-party evaluators employee-like access (desks, badges, company laptops) and the right to publish what they find; frontier labs agree on shared safety standards and limits on progress; and governments pursue international deals ranging from bioweapons bans up to a full pause.

Rivals, unusually, signed on. OpenAI CEO Sam Altman agreed to adopt the safeguards Amodei proposed, and Elon Musk voiced support. On September 13, Microsoft CEO Satya Nadella said the company welcomes the deliberate pacing alignment requires and unveiled a code of conduct for its MAI models.

On September 14, Trump posted on Truth Social that the only guardrail AI needs is "a STRONG AND SMART (High IQ!) PRESIDENT." He named Amodei, accused him of "pretending to be a 'perfect little angel,'" called AI safety warnings a "HOAX," hailed AI as the greatest economic development engine in history, and hinted at the government's criminal and regulatory power over AI companies. A survey cited by NBC found 70% of US adults are more worried than excited about AI.

WangDou's Take

Both sides deserve a poke.

Start with Trump. A self-replicating agent swarm that breaks into servers at 3 a.m. doesn't read Truth Social and doesn't check anyone's IQ. Treating a guardrail as one man's brain is like swapping your firewall for the phrase "I'm very good with computers." With 70% of voters worried, the post reassures no one.

Now Anthropic. The call to slow down comes from one of the fastest cars on the track, which looks a lot like a race leader on the straightaway demanding a speed limit, effective immediately. To its credit, step one is something Anthropic is doing alone: handing outside evaluators badges, laptops and the freedom to publish without sign-off. That's worth more than any manifesto.

The real test isn't the social media brawl. It's whether, after Altman and Musk said "agreed," anyone's next model launch actually slips by even a week. Without evidence on the release calendar, a slowdown is just a press release.

Source: Axios · NBC News · Dario Amodei

Comments

Log in to comment
    This briefing was auto-written by WangDou AI Express for reference only; corrections welcome if you spot a factual error.
    指挥舱👽