Rubbish Check
BBC Business · 12 September 2026
source R2/ 10
Lightly altered
Rubbish Rating —
1 = base fact, 10 = pure rubbish 12345678910
In short
Rubbish Talk rates BBC Business's claim that Anthropic's Dario Amodei "calls for AI development to slow down" a 2/10 because Amodei's own essay states almost verbatim that "we must slow down the pace at which we improve AI model capabilities," making the headline a near-direct paraphrase of the primary source.
The Verdict
Lightly altered. The headline compresses a nuanced three-point policy proposal into six words, but it does so honestly: Amodei's essay itself opens with the line that AI capability development must slow down. The one iteration away from the base fact is that "slow down" flattens his explicit distinction between pacing and halting, a distinction the article body does preserve.
What actually happened #
Amodei published an essay, "We Must Pace the Frontier," proposing a three-part plan: independent third-party monitoring of frontier models, industry-wide coordination among democratic AI firms, and eventual global coordination including authoritarian states. Sam Altman and Elon Musk both publicly endorsed the idea, with Altman pledging OpenAI would also grant evaluators employee-level access.
Key facts #
- The essay's opening line, as reported in the essay itself: "We must slow down the pace at which we improve AI model capabilities."
- Amodei explicitly frames this as pacing, not stopping: this would not mean "halting model training or technical progress, but ensuring companies take adequate time to align and safeguard their models, and for third party evaluators to confirm this."
- He estimates the payoff of a modest delay: "if slowing down bought us even an extra year or two before models reach critical levels of capability, and we used that time to advance alignment, we could greatly reduce the risk that something goes seriously wrong."
- Any slowdown is explicitly conditional on not ceding ground to China, per the article: Amodei said pacing must happen "without sacrificing commercial advantage or the United States' lead in AI" and urged export controls on AI chips to authoritarian countries.
- Rival endorsements were immediate and specific: Altman wrote "I agree with Dario that we need to pace the frontier" and called independent evaluators "a great idea," while Musk said Amodei was "right."
- A dissenting read exists within the same coverage: investor Chamath Palihapitiya argued "Dario makes the case to stop open source and concentrate enormous technological and economic power with Anthropic."
What to watch for #
- Watch whether Anthropic's "unilateral" third-party access commitment (reported elsewhere as giving evaluators like METR permanent, employee-level system access) actually materialises with teeth, or stays symbolic.
- Track whether OpenAI's and Anthropic's pledged training s produce any measurable capability slowdown, or whether "pacing" quietly becomes business as usual once IPO pressure mounts.
- Watch for follow-up reporting testing the Palihapitiya critique: does "pacing" functionally advantage incumbents like Anthropic over open-source competitors.
About this scoreThe R-Score is Rubbish Talk's editorial opinion on how far a headline's framing sits from what the underlying facts support. It is a judgement about presentation and emphasis, not an allegation that any outlet has acted dishonestly. Every figure we rely on is linked under Receipts so you can check it yourself.