US government orders first-ever restriction of AI model over jailbreak vulnerability

        June 18, 2026

US government orders first-ever restriction of AI model over jailbreak vulnerability

Also: US government orders Anthropic to shut down frontier models Fable 5 and Mythos 5 via export controls‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 

                X-Risk Daily

                Thursday 18 June 2026

Transformative AI

US government orders first-ever restriction of AI model over jailbreak vulnerability

On 12 June, the US government ordered Anthropic to disable Claude Fable 5 and Mythos 5 worldwide, just three days after Fable 5's 9 June release.

            Also in this issue

Transformative AI

US government orders Anthropic to shut down frontier models Fable 5 and Mythos 5 via export controls

Transformative AI

Trump executive order requires 30-day pre-release model submission to national security agencies

Transformative AI

NSA reportedly using Anthropic's Mythos model for offensive cyber operations

Geopolitics & Conflict

US and Iran sign nuclear agreement with $300bn redevelopment package

Research

Frontier AI models can detect when their responses have been tampered with, potentially undermining safety evaluations

Research

Researcher proposes 'overparameterization' theory to explain human-AI generalisation gap

            … and 44 more in the full briefing

Read today's briefing →

            Prefer a weekly digest? Click ‘manage your subscription’ below.

            Got thoughts on today’s briefing? Just reply to this email — I’ll read and reply!

            Generated automatically from 11 sources including Transformer, Sentinel,
            Epoch AI, LessWrong, and EA Forum.

                                Don't miss what's next. Subscribe to X-Risk Daily:

                        Name (required)

                        Frequency

            Email address (required)