The Information Machine

AI Labs Simultaneously Acknowledge Recursive Self-Improvement Threshold · history

Version 15

2026-06-25 18:23 UTC · 293 items

What

In June 2026, Anthropic and OpenAI independently disclosed data they described as early signs of recursive self-improvement in deployed AI systems and called for international coordination to slow frontier AI development [1][3]. The US government responded on June 12-13 by directing Anthropic to cut SK Telecom's access to Claude Mythos, citing alleged Chinese ties at the Korean carrier; Anthropic, lacking customer-level access controls, disabled both Fable 5 and Mythos 5 globally [14][15][17]. The models remain offline as of June 25, with some accounts suggesting restoration may come once the SKT access issue is resolved [20]. Social media attention to the shutdown has grown, with some observers calling it the most consequential AI regulatory development of 2026 [27].

Why it matters

The export control action establishes that US national security law can compel AI labs to revoke specific customer access, with global model shutdowns as the enforcement mechanism when targeted controls don't exist. The China-ties rationale shows that, at present, the government's primary concern with frontier AI is geopolitical rather than capability-based — even as labs frame the moment publicly as a genuine safety threshold.

Open questions

  • Are the alleged Chinese ties at SK Telecom verified, and has SKT responded publicly to that characterization? [14][15]

  • Will Fable 5 and Mythos 5 be restored once SKT access is severed, and on what official timeline? [20]

  • Some accounts reference a 'technical disagreement' as part of the shutdown rationale alongside the SKT access issue — what specifically does this refer to? [28]

  • Anthropic has not responded to Jeremy Howard's argument that its continued use of its top-ranked model for frontier research negates its slowdown call [24]; does the government-forced suspension change that position?

Narrative

In June 2026, Anthropic and OpenAI independently published documents acknowledging that recursive self-improvement may already be underway in deployed AI systems. Anthropic's 'When AI Builds Itself' disclosed that Claude authored more than 80% of Anthropic's production code merged in May 2026, that per-engineer code output reached 8x the 2024 baseline, and that Claude Mythos Preview accelerated model-training code approximately 52x [1][2]. Claude's open-ended coding success rate reached 76%, up 50 points in six months. OpenAI's concurrent policy blueprint called RSI 'potentially the most consequential frontier safety issue of the coming decade' and cited 'early signs of recursive self-improvement in today's systems' [3]. All three major labs publicly endorsed international coordination to slow frontier AI development [4].

Each lab's public safety stance sits in tension with other conduct. Anthropic's global slowdown call was timed to a confidential S-1 filing at roughly $965 billion [5][6], and Anthropic engineers were reported embedded at the NSA for offensive cyber using Claude Mythos [7]. Sam Altman publicly endorsed coordinated slowdown mechanisms [8] while telling staff that a major RSI breakthrough would favor staying private, since some work is easier without public-market pressure [9][10]; OpenAI simultaneously filed confidentially for an IPO [11]. The US government directed CAISI — the federal oversight body OpenAI itself proposed — to stop publishing public AI model evaluations, removing the transparency mechanism the proposal depended on [12][13].

The most concrete government action came on June 12, 2026, when the White House told Anthropic to cut SK Telecom off from Claude Mythos over alleged Chinese ties at the Korean carrier [14][15]. Anthropic, unable to restrict access at the customer level without taking the models fully offline, disabled both Fable 5 and Mythos 5 globally on June 13 [16][17]. SK Telecom had joined Project Glasswing — an Anthropic program focused on AI-assisted software vulnerability detection — and used a gated version of Claude Mythos to protect its core infrastructure [18][19]. Some accounts suggest the models may be restored once the customer access issue is resolved [20], though no official timeline has been confirmed. Claude Mythos 5 has been described as the first public test case for frontier-model export controls [21].

The governance and safety research response runs in parallel. Jack Clark's Import AI declared 'alignment is not on track,' arguing that RSI without adequate alignment means 'rolling very scary dice' [22]. Geoffrey Irving founded Sequent, backed by researchers from the UK AI Security Institute and Timaeus, on the argument that current lab alignment approaches are too empirical and reactive to provide principled safety confidence before ASI; the company plans to raise $100-150M and scale to 40-80 employees [23][22]. Jeremy Howard argued that the only internally consistent form of Anthropic's slowdown call would prohibit the lab with the top-ranked model from using it for frontier research — a test Anthropic fails — and Anthropic has not responded [24]. Dario Amodei separately disclosed that internal testers of Anthropic's most powerful unreleased model recommended against releasing it, describing it as a 'super weapon' — a characterization later clarified as a metaphor from early testers, not a literal licensing requirement [25][26].

Timeline

  • 2026-05: Claude authors 80%+ of Anthropic's production code; per-engineer output reaches 8x the 2024 baseline; Claude Mythos Preview accelerates model-training code ~52x; coding success rate reaches 76%, up 50 points in six months. [2][39][1]
  • 2026-06-03: OpenAI publishes 'Democratic Governance of Frontier AI,' acknowledging early RSI signs and proposing CAISI as a federal oversight body with mandatory evaluation authority. [12][3][35]
  • 2026-06-04: Anthropic publishes 'When AI Builds Itself,' disclosing Claude's 80%+ code authorship and calling for a global coordinated slowdown in frontier AI development. [29][40][41]
  • 2026-06-05: Critics note timing of Anthropic's slowdown call relative to its confidential S-1 filing at roughly $965 billion valuation. [5][40][6]
  • 2026-06-08: Sam Altman predicts AI will conduct a significant fraction of OpenAI's research by March 2028; OpenAI confidentially files for an IPO. [31][11]
  • 2026-06-09: All three major labs publicly endorse international coordination to slow frontier AI; NSA reported using Claude Mythos for offensive cyber with approximately six embedded Anthropic engineers. [4][7]
  • 2026-06-10: Geoffrey Irving founds Sequent; Jeremy Howard argues Anthropic's use of its own frontier models negates its slowdown call; Altman tells staff a major RSI breakthrough would favor staying private over pursuing the IPO. [23][24][9][10]
  • 2026-06-11: US government directs CAISI to stop publishing public AI model evaluations; Amodei tells Bloomberg AI progress is in the sharp-acceleration phase of an exponential. [13][30]
  • 2026-06-12: Google DeepMind identifies four distinct technical pathways through which AGI could transition to ASI. [33]
  • 2026-06-12: US government issues export control directive requiring Anthropic to block SK Telecom's access to Claude Mythos; SKT's Project Glasswing participation identified as the trigger. [42][17][43]
  • 2026-06-13: Anthropic disables Fable 5 and Mythos 5 globally for all customers, lacking customer-level access controls. [16][17][44]
  • 2026-06-15: Import AI 461 declares 'alignment is not on track'; Sequent confirms plans to raise $100-150M and scale to 40-80 employees. [22]
  • 2026-06-18: Reporting identifies the specific US concern as alleged Chinese ties at SK Telecom; White House directed Anthropic to cut SKT off from Claude Mythos specifically. [14][15][45]
  • 2026-06-19: Dario Amodei discloses testers of Anthropic's most powerful unreleased model recommended against releasing it; the 'super weapon' description clarified as a metaphor from early testers. [25][26]
  • 2026-06-20: Signal emerges that Fable 5 and Mythos 5 will become available again 'in the coming days'; shutdown enters day 10 with the SKT/China angle as the dominant public explanation. [20][46]
  • 2026-06-21: CNN characterizes AI regulation as 'a mess' with Anthropic 'caught in the crosshairs'; Claude Mythos 5 described as the first public test case for frontier-model export controls. [47][21]
  • 2026-06-25: Social media commentary continues to grow around the Fable 5/Mythos 5 shutdown, with some observers calling it the most consequential AI story of 2026 and underreported. [27][48][38]

Perspectives

Anthropic / Dario Amodei

Believes current models may be approaching the RSI threshold, calls for a global coordinated slowdown, and disclosed that testers of its most powerful unreleased model recommended against releasing it; Amodei says AI progress is in the sharp-acceleration phase of an exponential.

Evolution: The government-forced shutdown adds enforcement context but Anthropic has still not responded to Howard's consistency challenge or addressed the NSA offensive cyber reporting.

[29][1][7][30][25][17][14]

OpenAI / Sam Altman

Acknowledges RSI as the top frontier safety issue; official blog endorses coordinated slowdown mechanisms; Altman targets AI-conducted research at OpenAI by March 2028 and expects IPO within a year, but told staff a major RSI breakthrough might favor staying private.

Evolution: IPO filing and conditional private framing remain in tension with the public slowdown endorsement; CAISI's suppression of evaluations undercuts the transparency mechanism in OpenAI's own proposal.

[3][31][8][13][9][10][11][32]

Google DeepMind

Supports an international organization to enable coordinated slowdowns; researchers published technical work identifying four pathways from AGI to ASI.

Evolution: Governance position consistent with prior reporting.

[4][33]

Jack Clark (Import AI)

Declares 'alignment is not on track' and argues RSI without adequate alignment means 'rolling very scary dice'; treats the Fable/Mythos shutdown as confirming the risks of centralized AI control.

Evolution: Stance has sharpened from observational to declarative; the centralization-risk framing has gained traction following the model suspension.

[34][22]

Jeremy Howard

Argues the only internally consistent form of Anthropic's slowdown call would prohibit the lab with the top-ranked model from using it for frontier research; Anthropic fails this test and has not responded.

Evolution: Argument stands uncontested.

[24]

Geoffrey Irving / Sequent

Founded Sequent on the argument that current lab alignment approaches are too empirical and reactive to provide principled safety confidence before ASI; plans $100-150M raise and 40-80 researchers.

Evolution: Funding targets and research scope confirmed; position unchanged.

[23][22]

Zvi Mowshowitz

Guardedly positive on OpenAI's blueprint but warns federal preemption of state safety laws is its most dangerous element; treats CAISI evaluation suppression as a significant transparency setback.

Evolution: Consistent; concerned governance mechanisms are losing enforcement teeth.

[35][4][13]

Critics and observers (social media, decentralized-AI commenters)

Argue lab slowdown calls are competitive strategy timed to entrench incumbents; the Fable 5/Mythos 5 suspension has prompted growing observation that centralized AI can be shut off overnight; some observers describe it as the most important AI story of 2026 and underreported.

Evolution: Social media attention to the shutdown has grown through June 25; the 'incumbent entrenchment' and 'centralization risk' frames continue to circulate without new substantive evidence.

[5][6][36][37][27][38]

Tensions

  • Howard argues Anthropic's continued use of its top-ranked model for frontier research negates its slowdown call; Anthropic has not responded. [24]
  • Altman publicly endorses coordinated slowdown mechanisms while telling staff a major RSI breakthrough would favor staying private; the two positions point in different directions on what OpenAI actually wants from RSI timing. [8][9][10]
  • OpenAI proposed CAISI as a federal body with public evaluation authority; the US government directed CAISI to stop publishing evaluations, removing the transparency mechanism OpenAI's own proposal depended on. [12][13]
  • The US simultaneously embeds Anthropic engineers at the NSA for offensive cyber and issued export controls forcing Anthropic's most capable models offline globally; the China-ties rationale reveals the government's concern was geopolitical, not capability-based. [7][17][14]
  • Anthropic frames its slowdown call as a genuine safety response; critics argue it is regulatory strategy timed to a confidential S-1 at roughly $965 billion designed to entrench the incumbent's market position. [29][5][6]
  • Anthropic withheld its most powerful unreleased model after testers called it a 'super weapon,' while simultaneously deploying Fable 5 and Mythos 5 commercially until government action forced them offline — suggesting the internal threshold for withholding differs from what Anthropic considered safe for public deployment. [25][16]

Sources

  1. [1] Today’s edition of my newsletter just went out. — Rohan Paul Twitter (2026-06-05)
  2. [2] 😺 Anthropic: AI Is Building AI now — The Neuron (2026-06-05)
  3. [3] Peter Wildeford🇺🇸🚀 on X: "OPENAI: "We also see early signs of recursive self-improvement in today's systems". RSI is "potentially the most consequential frontier safety issue of the coming decade."" / X — reactive:rsi-governance-moment
  4. [4] Three Labs With a Plan and A Memorandum — Zvi's AI Roundups (2026-06-09)
  5. [5] The company that just confidentially filed its S-1 for a trillion-dollar IPO published a blog post four days later askin... — reactive:rsi-governance-moment (2026-06-05)
  6. [6] Anthropic has discovered the perfect pre-IPO narrative: its product is so powerful that it justifies a trillion-dollar v... — reactive:rsi-governance-moment (2026-06-07)
  7. [7] National Security Presidential Memorandum/NSPM-11 — reactive:rsi-governance-moment
  8. [8] OpenAI's latest official blog says the world may need a way to coordinate "slowing frontier development when needed." ht… — Rohan Paul Twitter (2026-06-09)
  9. [9] Sam Altman tells OpenAI staff an IPO is planned next year, but recursive self-improvement would favor staying private · Digg — reactive:rsi-governance-moment
  10. [10] Sam Altman tells OpenAI staff an IPO is planned next year, but recursive self-improvement would favor staying private · Digg — reactive:rsi-governance-moment
  11. [11] @skaas777 @iamai_omni 不是终止IPO啦，OpenAI 6月8日刚 confidentially filed for IPO，只是 timing 还没定（他们自己说有些事 private 公司做更方便）。 — reactive:rsi-governance-moment (2026-06-11)
  12. [12] [PDF] Democratic Governance of Frontier AI - OpenAI — reactive:rsi-governance-moment
  13. [13] AI #172: The First Fable — Zvi's AI Roundups (2026-06-11)
  14. [14] Alleged China ties at SK Telecom alarmed US officials and triggered Anthropic crisis — reactive:rsi-governance-moment
  15. [15] In early June, the White House told Anthropic to cut SK Telecom off from Claude Mythos over alleged Chinese ties, Wired ... — reactive:rsi-governance-moment (2026-06-18)
  16. [16] Anthropic disabled its two most advanced models, Fable 5 and Mythos 5, globally for all customers on June 13, 2026, afte... — reactive:rsi-governance-moment (2026-06-15)
  17. [17] Anthropic to disable its most advanced AI models after US order ... — reactive:rsi-governance-moment
  18. [18] SKT enlists gated Claude Mythos AI to shield core infrastructure — reactive:rsi-governance-moment
  19. [19] SK Telecom Joins Anthropic’s ‘Project Glasswing,’ Gains Access to Cybersecurity AI ‘Claude Mythos’ < Telecom < IT·Gaming < 기사본문 - The Elec Inc. — reactive:rsi-governance-moment
  20. [20] “In the coming days, [Fable 5 and Mythos 5] will become available again.” — reactive:rsi-governance-moment (2026-06-18)
  21. [21] Claude Mythos 5 has become the first public test case for frontier-model export controls. The question is not simply “wi... — reactive:europe-ai-sovereignty-deficit (2026-06-21)
  22. [22] Import AI 461: "Alignment is not on track"; FrontierCode; and synthetic research interns — Import AI (2026-06-15)
  23. [23] Sequent: scale and automation for higher confidence in alignment — Alignment Forum (2026-06-10)
  24. [24] Quoting Jeremy Howard — Simon Willison (2026-06-10)
  25. [25] Anthropic's CEO just went on record saying the people who tested their most powerful AI model came back asking them not … — Milk Road AI Twitter (2026-06-19)
  26. [26] @TannerLeidy @Polymarket Mythos doesn't literally require a gun license—it's a metaphor from early testers. — reactive:rsi-governance-moment (2026-06-17)
  27. [27] the Fable 5 ban is the most important AI story of 2026 and it's being underread. — reactive:rsi-governance-moment (2026-06-25)
  28. [28] Fable and Mythos were not just shut down because of a jailbreak claim. They were shut down because a technical disagreem... — reactive:rsi-governance-moment (2026-06-17)
  29. [29] Anthropic just called for a global way to slow frontier AI because its own models may be approaching recursive self-impr… — Rohan Paul Twitter (2026-06-05)
  30. [30] Dario Amodei's new interview, says AI progress suddenly going crazy. — Rohan Paul Twitter (2026-06-11)
  31. [31] Sam Altman's new blog about OpenAI's future path says by March-2028 a significant fraction of its own research will be d… — Rohan Paul Twitter (2026-06-08)
  32. [32] OpenAI expects to go public 'within the next year,' the Information reports — reactive:rsi-governance-moment
  33. [33] Beautiful paper from Google DeepMind. — Rohan Paul Twitter (2026-06-12)
  34. [34] Import AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racing — Import AI (2026-06-08)
  35. [35] OpenAI Offers A New Policy Blueprint — Zvi's AI Roundups (2026-06-05)
  36. [36] 👽 bittensor:native IS EXPLODING BECAUSE CENTRALIZED AI JUST PROVED IT CAN BE SHUT OFF OVERNIGHT — reactive:rsi-governance-moment (2026-06-16)
  37. [37] The real AI infrastructure story this week isn't a new model. It's the split between models you rent and systems you can... — reactive:rsi-governance-moment (2026-06-15)
  38. [38] The most consequential AI regulatory standoff of 2026 just entered its ninth day - and the underlying conflict reveals a... — reactive:rsi-governance-moment (2026-06-22)
  39. [39] Anthropic just disclosed that Claude now writes more than 80% of the production code it merges. — Rohan Paul Twitter (2026-06-05)
  40. [40] Anthropic calls for global AI slowdown after $965B valuation. Critics claim it's just to hobble competition. — reactive:rsi-governance-moment
  41. [41] Anthropic calls for pause of global AI development — reactive:rsi-governance-moment
  42. [42] In a US government export control directive issued June 12, 2026 ... — reactive:rsi-governance-moment
  43. [43] SK Telecom has been identified as the South Korean company whose inclusion in Anthropic's Project Glasswing programme tr... — reactive:rsi-governance-moment (2026-06-19)
  44. [44] Anthropic suspends top AI models after U.S. export control order — reactive:claude-fable-5-mythos-launch
  45. [45] 🚨BREAKING: SK Telecom Triggered Anthropic's Export Controls. Not Amazon — reactive:rsi-governance-moment (2026-06-18)
  46. [46] 🔍 Day 10 of the Fable 5 shutdown, and the real reason has finally come into focus. It is messier than the early "jailbre... — reactive:rsi-governance-moment (2026-06-22)
  47. [47] AI regulation is a mess, and Anthropic is caught in the crosshairs — reactive:rsi-governance-moment
  48. [48] Anthropic launched Claude Fable 5 on June 9. Three days later, the US government issued an export control directive and ... — reactive:rsi-governance-moment (2026-06-25)