G7 Summit Tackles AI, Ukraine and Children's Online Safety

Published 3 days ago

story

Above: G7 Leaders in Evian-les-Bains on June 16. Image credit: Evelyn Hockstein-Pool/Getty Images

The Spin

The government's export control on Anthropic's Fable 5 is a reckless overreach that punishes defenders more than bad actors. The jailbreak cited as justification is narrow, non-universal and replicable on GPT-5.5 and Chinese models alike — pulling best-in-class tools from security teams while adversaries advance freely is a strategic blunder. Transparent, science-based regulation is the only legitimate path forward.

FreeFable Anthropic

Anthropic championed Mythos as a cyberweapon requiring strict guardrails, then refused to patch a confirmed jailbreak when the government asked — that's a stunning betrayal of its own safety-first brand. National security has to outrank revenue cycles and pre-IPO valuations, full stop. The export control was a reluctant but necessary response to Anthropic's refusal to cooperate with a reasonable safety request.

DoW CIO Kirsten Davies on X

David Sacks on X

Metaculus Prediction

There is a 36% chance that, before 2029, a new international organization focused on AI safety will be established with participation from at least three G7 countries, according to the Metaculus prediction community.