Agency Says Frontier AI Risks Emerge Before Release as Safety Testing Falls Behind. The United Kingdom’s AI Safety Institute believes frontier AI models pose risks before they’re released to the public, and it doubled down on the need for voluntary safety commitments from AI companies.
First seen on govinfosecurity.com
Jump to article: www.govinfosecurity.com/uk-ai-safety-institute-warns-risks-in-unreleased-models-a-33078
