:
[AI]■ BREAKING

US EXPORT CONTROLS FORCE ANTHROPIC TO DISABLE ADVANCED MODELS

AI DESK1 MIN READ
FRI, JUL 24, 2026

■ AI-SUMMARIZED FROM 5 SOURCES ▸ TIMELINE

Anthropic has disabled its most advanced AI models globally after a US government export control directive. The move came following discussions between Amazon CEO Andy Jassy and US officials over national security concerns.

The US government ordered Anthropic to suspend access to Fable 5 and Mythos 5 for all foreign nationals, citing concerns that safeguards could be bypassed to identify software vulnerabilities. Rather than restrict access by geography, Anthropic chose to disable the models entirely. Reports indicate Amazon CEO Andy Jassy's conversations with US officials about the models' capabilities preceded the crackdown. The government did not provide Anthropic with specific details justifying the order. The decision has intensified international concerns about US dominance over AI development. Politicians and technologists outside the US view the move as confirmation that access to advanced AI systems will remain controlled by American interests. The directive represents one of the most significant AI export restrictions imposed by the US government to date, reflecting escalating tensions between AI safety, national security, and global technology access.

■ MORE FROM THE AI DESK

Israel and the UK have named officials to lead their artificial intelligence strategies, joining a growing list of nations establishing dedicated AI leadership roles. The appointments reflect intensifying global competition to establish AI dominance.

1H AGOAI Desk

Black Forest Labs released Flux 3, advancing open-source image generation with improved quality and speed. The model demonstrates significant progress over previous versions.

1H AGOIndustry Desk

Centrica, owner of British Gas, is eliminating 1,300 call centre positions while deploying AI tools, with the CEO claiming customers prefer chatbots to human staff.

3H AGOAI Desk

Apple's updated Siri successfully executes multistep prompts and understands context well, according to hands-on testing by The Verge. The assistant features robust guardrails but operates with less personality than Google's Gemini.

3H AGOAI Desk

■ SUBSCRIBE TO THE DAILY BRIEF

ONE EMAIL, 5 STORIES, 06:00 UTC. UNSUBSCRIBE ANYTIME.