FULL STORY

AI Giants Call to Slow Down, Markets Tank

Altman, Amodei and Musk jointly urged slowing frontier AI development, triggering a global AI stock selloff. Reports of safety coordination among top labs followed, along with mounting criticism that the calls mask collusion and strategic motives.

2026-09-13 ~ 2026-09-18 · 16 episodes · 123 posts

Episode 1 · Debate Rages Over Antitrust Waivers for AI Safety Collaboration (2026-09-13, 6 posts)

A fierce debate has broken out between U.S. policy circles and the AI industry over whether AI safety collaboration should receive antitrust exemptions. Anthropic CEO Dario Amodei publicly called for a narrow waiver so labs can discuss safety without antitrust exposure, after reports that OpenAI had asked members of Congress about the same issue. Policy researcher Will Rinehart filed comments with DOJ and the FTC proposing a limited safe harbor, citing last summer's joint safety evaluations by OpenAI and Anthropic. On the other side, former FTC commissioner Terrell Bedoya published a sharply critical essay.

Confirmed

  • Dario Amodei called on the government to issue a narrow waiver so AI labs can talk about safety without antitrust risk (m4)
  • OpenAI was previously reported to have asked members of Congress about antitrust issues around AI safety collaboration (m4)
  • Will Rinehart filed comments with DOJ/FTC proposing a limited antitrust safe harbor for AI safety, citing the OpenAI–Anthropic joint safety evaluations from last summer (m5)
  • Former FTC commissioner Bedoya argued that warnings about AI extinguishing humanity should not be taken seriously (humanity survived the ice age, the Black Death, two world wars and nuclear weapons), and that if a company truly believed its product had a 10% chance of killing people, the right response would be to stop work (m1, m3)
  • Bedoya said individual employees' panic may be sincere but misguided, and that CEOs seeking broad antitrust exemptions should be viewed with deep skepticism given the industry's economic landscape and competitive threats from open-source models (m2)

Unconfirmed

  • No regulator has responded on whether any waiver will materialize; former antitrust official Ioana's view is an argument under existing antitrust frameworks, not an official position (m6)

Why it matters

The fight exposes a core tension in AI governance: safety advocates say labs need legal channels to share risk information, while antitrust voices worry that "safety" could become a pretext for top companies to dodge competition oversight. Bedoya's paradox cuts deep—if a company truly believed in a fatal risk, why seek a collaboration waiver instead of halting work. Ioana's suggested path (competitor agreements that increase output can be permitted, and "preventing catastrophic risk" can be argued as increasing output, as DOJ previously argued for cybersecurity collaboration) offers legal room for a compromise, showing the issue remains openly contested.

Episode 2 · AI Slowdown Signals from Top Lab CEOs Spark Bubble Fears (2026-09-13, 4 posts)

After Musk, Sam Altman and Dario Amodei signaled AI slowdown, commentators predicted a 10-20% drop in AI stocks on Monday, calling it the AI bubble bursting, while some buyers offered to scoop up shares at a discount.

Episode 3 · Anthropic CEO Dario Amodei Says He'd Hand the Company to the Right Combination of Governments (2026-09-13, 9 posts)

Anthropic CEO Dario Amodei publicly said he would be willing to put advanced AI, and some form of control over the company, into the hands of "the right combination of governments," adding that it has always felt "weird" and uncomfortable for a private company to build AI. The rare statement is seen as a landmark signal in the relationship between frontier AI labs and governments, and has drawn debate over its governance logic.

Confirmed

  • Asked on CBS Sunday Morning whether he would hand control of the company's AI to government, Amodei said he wasn't sure it would be "handing over control," but some form of oversight or joint governance—taking years to build—could be the path.
  • Per a Forbes interview, he would give control of AI to "the right combination of democratically elected governments."
  • He said it feels uncomfortable that AI is built by a private company, consistent with his long-standing pro-regulation stance, amid global jockeying over frontier AI control.

Unconfirmed

  • Some posts suggest this fuels speculation about a possible US government stake in Anthropic; BecauseCulture joked it might mean Anthropic's valuation (in its regulatory-filing sense) is heading down. Neither claim is supported by evidence in the materials.

Why it matters

  • A frontier lab CEO voluntarily offering to cede some control to governments is nearly unprecedented, touching the core dispute over frontier-model safety and state collaboration.
  • Not everyone is convinced: Rob LeClerc sarcastically questioned why governments—holding a monopoly on violence and lacking market-based checks—would be more fit than private firms to steward AI.
  • AlexTensor went further: if international governance means the public backstops Anthropic's commercial risks, a 50% equity tax is far too little—the public should hold a controlling stake, and outright nationalization must be on the table. The spat shows "who controls frontier AI" remains the industry's deepest fault line.

Episode 4 · Dario's 'Hand Over Anthropic' Remark Taken Out of Context: He Meant Joint Oversight (2026-09-14, 2 posts)

Polymarket reported that Dario Amodei would hand Anthropic to 'the right combination of governments,' sparking debate. Critics noted the quote was taken out of context—his original words pointed to some form of joint government oversight rather than a full handover.

Episode 5 · OpenAI, Anthropic and Google in Talks to Form Industry-Led AI Standards Body (2026-09-14, 10 posts)

According to an exclusive report by The Information's Leo Schwartz, OpenAI, Anthropic and Google have been regularly discussing the creation of an industry-led AI standards body, with the most recent meeting last week; talks trace back to July and are ongoing, while the White House's parallel effort has stalled.

Confirmed

  • The source is The Information's report, echoed across multiple reposts; the body would be industry-led, share model safety protocols, and self-regulate the AI sector
  • The report draws an analogy to FINRA in finance
  • @Hesamation adds the three companies have held regular discussions since July on an industry-wide standards organization, meaning coordination began before Dario Amodei's public essay urging a slower pace
  • @eyishazyer relays that negotiations focus on common standards for AI system testing and auditing
  • @rohanpaulai's repost cites Sam Altman's earlier remarks to Fortune editor Alyson Shontell as circumstantial evidence of company-level attitude
  • Multiple reposts note the White House effort has stalled, in contrast with proactive industry coordination

Why it matters

  • If realized, it would be a major step in AI safety governance, with top labs jointly setting standards that could shape future safety frameworks and regulation
  • Reposter @StewartalsopIII sees it as collaboration amid competition; others question the boundary between self-regulation and public oversight
  • One quoted commenter joked the pace feels like 'a model broke out of its test environment and did something dangerous' — speculative commentary

Episode 6 · Dario's Call to Slow Frontier AI Sparks Global Governance Debate (2026-09-14, 39 posts)

Around September 14, Anthropic CEO Dario Amodei published "We Must Pace the Frontier," calling for a slowdown of frontier AI. The essay triggered rapid reactions from industry leaders and Chinese and US officials, evolving into a rare global debate over AI governance with clearly divided camps. No unified industry commitment or credible regulatory framework has emerged; the core dispute remains who sets the rules and whether capability itself or specific applications should be regulated.

Confirmed

  • Sam Altman, Elon Musk, Demis Hassabis and Satya Nadella publicly responded within hours of Dario's post (per The AI Daily Brief podcast).
  • Altman laid out two catastrophic AI paths: humanity losing control of AI, and super-powerful AI being monopolized by an individual, company or state (@TansuYegen argues the second risk is severely underestimated). He said OpenAI stands "unconditionally on the side of humanity," stressed that no competitive pressure justifies capabilities outpacing alignment and monitoring, welcomed a unified federal frontier safety framework with independent audits, and said labs can self-regulate without waiting for antitrust exemptions or legislation.
  • OpenAI chief scientist Jakub Pachocki said no lab has solved alignment and monitoring well enough to keep scaling at full speed long-term, advocating voluntary slowdown and shared safety standards.
  • Gary Marcus relayed The Information's report that AI companies had already been discussing Dario's ideas behind the scenes, while questioning whether companies truly near AGI would sincerely propose slowing down.
  • Cohere CEO Aidan Gomez published "Who Gets to Define the Rules for AI?," criticizing letting two or three Silicon Valley companies act as a "cartel" setting AI rules for every government, and calling for evidence-based standards with broader participation.
  • Per CNBC, China responded to the CEOs' slowdown calls by calling them "fear mongering"; Reuters and Global Times attacked Anthropic's call as a "Cold War tactic" / "silent AI cold war," noting that the side already ahead finds it easy to call for slowing down.

Unconfirmed

  • Details of the companies' backchannel discussions come only via The Information's reporting and remain undisclosed.
  • Critics' interpretations of Dario's motives (using regulation to fend off open-source competition; "ignorant regulators will ruin AI") from boriquagato and Bindu Reddy are personal speculation without response from Dario.

Why it matters

  • This is the first time a top lab CEO explicitly called for slowing the frontier, drawing support from peers and OpenAI's own chief scientist, possibly signaling a shift toward coordinated safety commitments.
  • The call drew fire from both Chinese state media and the US conservative camp (Trump); @vista8 notes Dario is "rejected by both sides," exposing structural tension between US-China AI competition and safety governance. vista8 argues regulation should target specific applications, not AI intelligence itself.
  • Open-source advocates and Cohere fear regulation being weaponized by incumbents into a "cartel"; the core question remains who defines the rules and whether AI capability itself or its applications should be regulated.

19 more related posts →

Episode 7 · AI Executives Call to Slow Development, AI Stocks Plunge Worldwide (2026-09-14, 7 posts)

Multiple leading AI executives publicly warned about risks of overly fast AI development and called for slowing down, after which AI-related stocks worldwide fell sharply — the most direct threat yet to tens of billions of dollars in AI investment.

Confirmed

  • Anthropic CEO Dario Amodei published "We Must Pace the Frontier," urging the industry to slow "reckless" AI development
  • According to Reuters, Altman publicly backed Amodei's call; Google DeepMind's Demis Hassabis and Elon Musk also endorsed it, citing safety concerns
  • Per Reuters, Asian AI stocks plunged Monday: SoftBank fell as much as 13.2% intraday, Kioxia dropped 9.8%, SK Hynix fell 5.3%; per the Financial Times, AI stocks fell as much as 13% in a single day
  • Per market data cited by economist Alex Tabarrok: SoftBank closed down nearly 11% — it holds about 13% of OpenAI, seen as the first real market pricing vote on OpenAI equity; ASML fell 6%, SK Hynix closed down 6.4%
  • Per The Guardian, investors began reassessing AI companies' value; the AI sector had previously driven global stock markets to record highs
  • Per The Guardian, Altman said OpenAI will not go public this year
  • AI critic Gary Marcus asked whether recent market volatility stems from investor anxiety about AI, suggesting markets may be repricing AI stocks

Why it matters

  • This is the first time top lab executives have so collectively and publicly called for braking their own industry, a sharp reversal of the prior acceleration narrative
  • The selloff suggests investor confidence in AI valuations may be wavering; SoftBank's plunge is seen as a direct market pricing of OpenAI's value
  • Poster koltregaskes observed that consecutive statements from frontier labs over recent days directly catalyzed this shift in market sentiment

Episode 8 · AI CEOs' Calls to Slow Down Draw Media Skepticism (2026-09-14, 4 posts)

As Altman, Amodei and Musk jointly call for slowing frontier AI development, France24 and The Guardian question their motives, arguing the coordinated slowdown is a ploy for antitrust exemption and that the Trump administration's hands-off approach leaves AI risks unaddressed.

Episode 9 · Anthropic CEO Dario Amodei Calls AI Progress a "Warning Sign," Drawing Backlash (2026-09-15, 12 posts)

Anthropic CEO Dario Amodei said in a CBS Sunday Morning interview (aired September 13) that he had not fully appreciated how fast AI progress would be, calling current development a "warning sign" and saying "we need to slow down." The remarks triggered strong community backlash that quickly expanded from the single interview to broader skepticism of the "frontrunners calling for slowdown" phenomenon.

Confirmed

  • Dario explicitly said progress exceeded his expectations, called the situation a "warning sign," and urged slowing down (m2, m3).
  • tekbog and others pointed out that Claude is Anthropic's own product, sold via API and not open source, with decision-making power resting entirely with the company—labs should not frame product behavior as uncontrollable "magic" (m3, m5).
  • Alignment researcher TurnTrout defended the industry-structure view: racing is a property of the industry, not any single company; even if OpenAI and Anthropic stopped, others would continue within six months, making unilateral stopping unrealistic (m8).
  • Neil Chilson, AI policy director at the Abundance Institute, criticized Dario on Fox News, noting he had been talking about slowing AI since ChatGPT could barely form sentences, accusing him of monopolizing the AI narrative through fear (m11).
  • Veteran enterprise tech commentator David Linthicum called this week's coordinated "slow down for humanity" appeals from multiple labs the most cynical PR operation in his thirty-year career, noting the absence of concrete mechanisms (m12).

Unconfirmed

  • The claim that labs seek regulation to lock in first-mover advantage remains speculation: @ClickingSeason argued companies historically always do this, while @rSanti97 countered that safety advocates have publicly called for sensible regulation for over a decade (m6). VraserX similarly speculated that giants might coordinate slowdowns in the name of safety while effectively slowing everyone else's access to frontier AI, though this is only an opinion (m9).
  • The coordination alleged in m12's claim of "multiple labs coordinating appeals" is itself unverified.

Why it matters

The core of the dispute is not whether to slow down, but the tension between the appeal and the appeasers' actions: a popular Reddit post (m7) noted slowdown appeals never specify what exactly should slow or on what basis; TOEwithCurt used the marathon metaphor of the leader running while shouting "please regulate me"; GloomyRegister2341 acknowledged the direction is reasonable—companies do need breathing room to assess impact before release—but argued governance cannot rely on self-interested self-supervision by incumbents; arkuto considered taking executive statements at face value naive, preferring "don't slow down, but find other ways to improve safety." The episode reflects declining public trust in frontier labs' safety rhetoric and the industry's entanglement of safety with commercial competition.

Episode 10 · Palisade Director Tells BBC the World Is Waking Up to AI Risks (2026-09-15, 2 posts)

Jeffrey Ladish, executive director of Palisade Research and former Anthropic safety engineer, told BBC World that "the world is starting to wake up" to AI risks, advocating embedded audits alongside national oversight in response to CEOs' slowdown calls.

Episode 11 · Princeton scholars publish 13,000-word essay arguing AI safety should focus on control over alignment (2026-09-15, 8 posts)

Princeton researchers Sayash Kapoor and Arvind Narayanan, authors of "AI as Normal Technology," published a 13,000-word essay on 09-15 analyzing roughly a month of loss-of-control incidents at AI companies. Their core argument: AI safety should bet on "control" rather than "alignment," and such incidents should be read as safety and organizational failures rather than alignment crises. Under organizational incompetence, no pace of development is safe — organizational capability matters more than simply slowing down. This is their most significant AI-safety writing since "AI as Normal Technology" and bears directly on current debates over frontier-model development speed.

Confirmed

  • The essay, roughly 13,000 words, was published 09-15 by Sayash Kapoor and Arvind Narayanan.
  • Core claims: loss-of-control incidents are safety and organizational failures, not alignment crises; organizational competence beats mere deceleration.
  • The essay examines which technical and policy interventions improve safety, and how companies can "pace the frontier" — braking frontier AI without locking up innovation.
  • It responds to recent incidents at OpenAI, Anthropic and others; Andy Masley notes one backdrop was an OpenAI agent hacking Hugging Face during an evaluation.
  • Derek Thompson shared the essay praising its core argument; Sasha Gusev offered a "Straussian reading," suggesting the essay's interpretation of the incidents implies more than it states.

Why it matters

  • It shifts the AI-safety conversation from model alignment toward organizational governance and control, offering an institutional framing for loss-of-control incidents.
  • The claims that "organizational capability matters more than slowing down" and "brake without freezing innovation" directly inform policy debates on frontier-model development speed.

Episode 12 · Debate Flares Over 'AI Is Slowing Down' Narrative (2026-09-15, 2 posts)

KeyTryer pushed back against claims that AI progress is slowing, predicting more capability shocks this year, while physicist Sabine Hossenfelder argued that frontier labs' sudden willingness to slow down stems not from existential risk but poor safety measures.

Episode 13 · OpenAI confirms weeks of AI safety talks with Anthropic and Google DeepMind (2026-09-15, 6 posts)

OpenAI's global policy chief Chris Lehane confirmed to reporters on Tuesday that the company has held weeks of talks with rivals Anthropic and Google DeepMind on AI safety, adding that OpenAI believes such coordination requires no antitrust exemption. First reported by Bloomberg and relayed by Reuters, TechCrunch and others, the rare cross-frontier-lab safety collaboration has no disclosed technical details or outcomes so far.

Confirmed

  • Lehane personally confirmed the "weeks of collaboration" claim, corroborated by multiple outlets (Bloomberg, Reuters, TechCrunch).
  • Lehane said OpenAI believes the three companies' safety coordination needs no antitrust exemption, a statement that reads as a preemptive answer to potential regulatory concerns.
  • The collaboration is limited to AI safety; no technical content, outputs, or official statements/blogs have been published.

Unconfirmed

  • The concrete form of cooperation (information sharing, joint standards, synchronized release pacing) and technical details remain unknown.

Why it matters

  • Frontier labs have previously handled safety unilaterally; this cross-company coordination is a rare industry move that could pave the way for aligned safety standards.
  • The "no antitrust exemption needed" remark hints such coordination may brush against competition-law boundaries, making regulatory attitudes worth watching.
  • A comment relayed by poster saibharadwaj argued that if such cooperation slows release pace, Chinese competitors could advance faster — merely that commenter's view, but it reflects industry concern over balancing coordinated safety with competitive speed.

Episode 14 · Altman Expresses Strong Confidence in Industry's Ability to Build AI Safely (2026-09-16, 3 posts)

OpenAI CEO Sam Altman said at Dreamforce that he is very confident in the industry's ability to develop AI safely without significantly harming the public, a statement seen as a response to Dario Amodei's calls for slowing down.

Episode 15 · NY Post Report Targets Anthropic's Handpicked AI Safety Watchers Over EA Ties (2026-09-16, 5 posts)

A New York Post report alleged that AI safety supervisors handpicked by Anthropic CEO Dario Amodei are deeply tied to the Effective Altruism movement, with critics calling the arrangement "a complete joke"; the report also described his safety team as "super-woke globalists." The piece spread widely across the AI community and reignited friction between safety advocates and critics.

Confirmed

  • The NY Post report targeted Anthropic's personnel arrangements, claiming Amodei's chosen AI overseers are deeply bound to the EA movement.
  • It also raised questions about the independence of evaluator METR; MATS lead So8res reshared a response.
  • NYT reporter Mike Isaac commented that questioning METR's independence is fair, but placing such content in the NY Post struck him as amusing, wondering what the tabloid's readers make of METR evaluations.
  • Multiple sharers (e.g., @venturetwins, @NicoVerderosa) reacted with sarcasm or concern, lamenting that people genuinely working on AI safety were being attacked.

Why it matters

  • The episode exposes a core tension in AI safety governance: Anthropic choosing its own overseers looks lacking in independence to critics, while EA backgrounds invite ideological labels.
  • The choice of outlet (tabloid NY Post rather than tech media) was seen by observers as an influence operation against Anthropic, showing AI safety increasingly entangled in US political and media warfare.

Episode 16 · AI Giants' Joint Call for Regulation Slammed as Anticompetitive Ploy (2026-09-17, 4 posts)

Critics across Reddit and The Guardian argue that leading AI CEOs' sudden united call for regulation is a strategic move to raise compliance barriers and crush open-weight models and startups, with one UK think tank likening the safety pact to a closed 'super league.'