US AI regulation fight: Amodei's safety coordination plan meets antitrust skepticism as Trump calls AI risk fears a 'hoax'

Anthropic CEO Dario Amodei proposed a three-step AI safety plan (third-party evaluators, industry coordination, international agreements) with tentative buy-in from Sam Altman and Demis Hassabis, while former DOJ antitrust chief Jonathan Kanter argued…

The Verge's continuing coverage of the US AI regulation fight (both reports dated 2026-09-19) pairs a safety-coordination pitch with a skeptical antitrust take. Anthropic CEO Dario Amodei proposed a three-step plan for slowing AI development: embedding third-party evaluators in labs, industry coordination, and international agreements; OpenAI's Sam Altman and Google DeepMind's Demis Hassabis tentatively agreed. According to a Wall Street Journal report cited by The Verge, Meta's Mark Zuckerberg, Elon Musk, and Nvidia's Jensen Huang scuttled a proposal for a FINRA-like, industry-funded AI regulator. President Trump dismissed AI safety fears as a 'hoax,' though critics noted his first administration signed Executive Order 13960 (2019) on federal AI use principles, which remains unrepealed, and that he retains criminal and regulatory power over AI companies. OpenAI's global affairs chief Chris Lehane urged Congress to adopt bipartisan, mandatory national safety standards for frontier AI. On the competition side, The Verge's Decoder podcast interviewed former DOJ Antitrust Division head Jonathan Kanter, who argued that frontier labs such as OpenAI, Anthropic, and Google DeepMind should not receive antitrust exemptions to coordinate on safety under either a generous interpretation (fear of unregulated risk) or a cynical one (regulatory capture, cartel formation, and IPO valuation protection). Kanter, citing Amodei's 'pace the frontier' framing, suspects the labs want permission to slow spending, protect IPO valuations, and foreclose open-weight Chinese competitors. He referenced his antitrust cases against Google, Apple, and Ticketmaster brought with Lina Khan; the Apple case continues under the Trump DOJ. The two threads connect directly: the industry safety coordination Amodei proposes is exactly the conduct Kanter says should not earn an antitrust exemption.

• Anthropic CEO Dario Amodei proposed a three-step plan for slowing AI development: third-party evaluators embedded in labs, industry coordination, and international agreements.
• OpenAI's Sam Altman and Google DeepMind's Demis Hassabis tentatively agreed with Amodei's plan.
• A Wall Street Journal report cited by The Verge says Meta's Mark Zuckerberg, Elon Musk, and Nvidia's Jensen Huang scuttled a proposal for a FINRA-like, industry-funded independent AI regulator.
• President Trump dismissed AI safety fears as a 'hoax,' but retains criminal and regulatory power over AI companies.
• Trump's 2019 Executive Order 13960 on federal AI use principles remains unrepealed, which critics cite as evidence of inconsistent policy.
• OpenAI global affairs chief Chris Lehane urged Congress to adopt bipartisan, mandatory national safety standards for frontier AI.
• Former DOJ Antitrust Division head Jonathan Kanter argued frontier labs should not receive an antitrust exemption to coordinate on safety under any interpretation.
• Kanter's cynical reading is that labs want regulatory capture, cartel formation, IPO valuation protection, and to foreclose open-weight Chinese competitors; the generous reading is fear of unregulated risk.