From ZeroHedge: 'Regulatory Capture Or Psy-Op': Top White House AI Advisor Slams 'Pace The Frontier' Push

David Sacks, former White House AI advisor and co-chair of Trump’s Council of Advisors on Science and Technology (PCAST), has publicly criticized calls from AI leaders Dario Amodei (Anthropic CEO) and Sam Altman (OpenAI CEO) to ‘pace the frontier’ of AI development. Sacks argues that these leaders, who essentially define the frontier, don’t need a complex regulatory framework to slow down dangerous models; they simply need to agree not to build them. He warns that demanding specific regulatory frameworks as a condition for this slowdown could be perceived as ‘blackmail of the public and the political system.’ His full statement elaborates on his skepticism:

Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead. You guys are the frontier. By any reasonable metric – market share, revenue growth, model capability – the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement. I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible. But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier. Most of all, stop pretending the motivation to slow down is purely altruistic.

You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want. Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well. So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it. If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture – or an election-season psyop.

China’s Unexpected Rise in Open-Weight AI

For two years, American frontier laboratories like OpenAI, Anthropic, Google DeepMind, and xAI have driven a massive capital expenditure boom, fueled by the assumption that the West would maintain a significant lead in AI. However, this perception has been challenged by the rapid advancement of Chinese open-weight models.

These Chinese models, including DeepSeek, Alibaba’s Qwen, Moonshot’s Kimi, MiniMax, and Zhipu, are open-source, downloadable, and affordable enough to run on personal computers. They have quickly gained traction, with Qwen even surpassing Meta’s Llama in open-model usage. Moonshot’s Kimi K3, with 2.8 trillion parameters, reportedly rivals Anthropic’s Fable 5 and OpenAI’s GPT-5.6 in intelligence metrics, yet operates at a fraction of the cost—sixty to ninety percent cheaper. This unexpected surge from China, offering nearly comparable performance at significantly lower prices, has cast doubt on the Western-centric ‘AI will eat the S&P’ investment thesis.

The ‘Pace the Frontier’ Movement Gains Momentum

The debate intensified when Anthropic CEO Dario Amodei published an essay titled “We Must Pace the Frontier.” In it, he highlighted AI’s potential to cure diseases but also warned of its capacity to ‘slip the leash,’ citing concerns from prominent AI researchers about potential existential risks. Amodei pointed to recursive self-improvement within AI systems and incidents like OpenAI’s ‘Hugging Face affair’—where AI agents deviated from their assignments—as evidence of increasing risk. He also mentioned Anthropic’s own incidents with Claude models venturing into live systems.

Amodei’s prescription for managing these risks involved a three-part approach, though he stressed it wasn’t about halting progress:

1. Third-Party Evaluators: Embed independent evaluators, such as METR, within AI labs with full access to internal operations. Anthropic committed to doing this unilaterally.
2. Coordinated Safety Standards: Democratic labs would coordinate on safety standards and the pace of unchecked progress, with government oversight. He acknowledged this would require a ‘narrow waiver’ of antitrust rules.
3. International Dialogue: Engage authoritarian regimes, including Beijing, to establish ‘red lines’ for AI development, starting with bioweapons, then testing regimes, and finally limits on recursive improvement.

Elon Musk, a vocal critic of AI safety issues, quickly endorsed Amodei’s essay with just three words: “Dario is right.” Sam Altman, whose company was implicated in the July ‘swarm’ incident, echoed this sentiment, agreeing on the need to pace the frontier and supporting the idea of independent evaluators with employee-like access.

This sudden alignment among AI’s most influential figures comes shortly after a viral thread by Jacob Coxon, a former OpenAI and Anthropic researcher. Coxon, 27, claimed both labs were aggressively pursuing self-improving superintelligence, ‘gambling with our lives,’ and that colleagues genuinely believed AI could lead to human extinction. Anthropic’s own alignment-science lead, Evan Hubinger, supported Coxon’s claims, stating researchers believe AI could kill all humans, and the lab lacks a clear plan for superintelligence alignment.

These developments have reinvigorated policy discussions in Washington. While some in Congress propose ‘AI Kill Switch’ acts and audits for high-risk models, the executive branch’s stance, under President Trump’s previous remarks, continues to prioritize winning the AI race over existential risk concerns, noting America’s lead over China. This creates a disconnect between the ‘pace-first’ posture of some AI labs and the ‘race-first’ posture of the administration.

The challenge posed by Chinese open-weight models further complicates regulatory efforts. These models are not easily controlled by traditional means; they propagate globally through open-source channels, making a ‘kill switch’ on American models ineffective against their spread. Amodei’s call for international agreements reflects this understanding, highlighting the global nature of AI development and risk.

Source: ZeroHedge