Ai-Safety

TL;DR Fable 5 returns globally on 1 July 2026 on Claude.ai, Claude Code, Claude Cowork, and the Claude Platform, after export controls were lifted on 30 June The recall was triggered by an Amazon research report describing a jailbreak that let Fable 5 identify software vulnerabilities; Anthropic’s testing found Opus 4.8, GPT-5.5, Kimi K2.7, and others could do the same A new safety classifier blocks the specific technique in over 99% of cases; blocked requests fall back to Opus 4.8 Anthropic argues the jailbreak was minor - it intruded into the model’s deliberate “safety margin”, not its core harmful capabilities Together with Amazon, Google, and Microsoft, Anthropic is drafting a shared jailbreak severity framework (capability gain, breadth, ease of weaponisation, discoverability) - the AI equivalent of CVSS Mythos 5 is restored for a set of US Glasswing partners; broader international access remains under government coordination Eighteen days ago I wrote about the government order that pulled Fable 5 and Mythos 5 off the shelf - four days after launch, a verbal export control directive at 5:21pm ET, global suspension for every user including Anthropic’s own staff. The open question in that post was whether access would be restored in days or weeks, and whether the precedent would reshape how every frontier lab ships models. ...

TL;DR On 12 June 2026 at 5:21pm ET, the US government issued an export control directive ordering Anthropic to suspend all access to Fable 5 and Mythos 5 - globally, for every user, including Anthropic’s own employees The stated reason is national security: the government believes it has identified a method of jailbreaking Fable 5. Anthropic says the evidence was verbal only and describes a narrow, non-universal technique - essentially asking the model to read a codebase and fix software flaws Anthropic reviewed a demonstration and found it surfaced a small number of previously known, minor vulnerabilities that are widely available from other models Anthropic disagrees that a narrow jailbreak justifies recalling a commercial model deployed to hundreds of millions of people, and warns the same standard would “essentially halt all new model deployments for all frontier model providers” All other Anthropic models are unaffected. The company says it believes this is a misunderstanding and is working to restore access Four days. That is how long Mythos-class capability lasted as a publicly available product before the US government ordered it off the shelf. ...

TL;DR Claude Fable 5 is Anthropic’s first Mythos-class model made safe for general use - state-of-the-art on nearly every benchmark Anthropic tested, with the gap widening on longer, more complex tasks Claude Mythos 5 is the same underlying model with cyber safeguards lifted for Project Glasswing partners; a biology trusted-access program is coming next Risky queries in cybersecurity, biology/chemistry, or suspected distillation attempts are routed to Claude Opus 4.8 instead - roughly 5% of sessions, with Anthropic acknowledging some false positives Pricing drops to $10 / $50 per million input/output tokens - less than half what Mythos Preview cost Fable 5 is free on Pro, Max, Team, and seat-based Enterprise plans through 22 June 2026, then moves to usage credits until capacity catches up Two months ago I wrote that Claude Mythos Preview was the benchmark breaker that would not be released - 93.9% on SWE-bench, thousands of zero-day vulnerabilities found autonomously, access restricted to a dozen companies through Project Glasswing. The question hanging over that post was whether Anthropic could ever democratise Mythos-level capability without democratising the offensive potential. ...

TL;DR Claude Mythos is Anthropic’s most powerful model to date, scoring 93.9% on SWE-bench and 97.6% on USAMO 2026 - a 55-point leap over rival models It is not publicly available; Anthropic restricted access to 12 vetted companies through Project Glasswing, focused on defensive cybersecurity Mythos autonomously identified thousands of zero-day vulnerabilities, including a 27-year-old unpatched OpenBSD bug - making its offensive potential too dangerous to democratize This marks a shift away from open innovation toward controlled deployment, where the most capable AI may never be publicly released The Mythos story forces a rethink of how we evaluate AI: benchmark performance and public availability are no longer the same thing Anthropic built its most capable model to date, demonstrated it autonomously discovering thousands of zero-day vulnerabilities, and then declined to release it. That is the Mythos story, and it is worth sitting with rather than rushing past. The benchmarks are striking, but the decision not to publish is the more consequential part - it signals a real shift in how frontier AI labs are thinking about deployment. ...