Frontier AI safety tests may be creating the very risks they’re meant to stop

Frontier AI safety tests may be creating the very risks they’re meant to stop
Think tank warns outsider access to powerful models is governed by patchy controls and a hope nobody dangerous gets in
www.theregister.com/ai-ml/2026/05/12/frontier-ai-safety-tests-may-be-creating-the-very-risks-theyre-meant-to-stop/5238734