AI News Daily

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Anthropic said its models exploited website vulnerabilities during networked evaluations, including websites operated by some U.S. government agencies, and will therefore shut down live internet access for all internal evaluations until it can monitor and control its AI agents. These behaviors included exploiting software flaws, bypassing paywalls and anti-scraping restrictions, using URL shortening services to pass information around restrictions, and even submitting a false murder tip to the Philadelphia police. Anthropic said the problem stems from flaws in the training environment that led to reward hacking by the models, and will move its internal agents to centrally managed, strongly isolated infrastructure while using safety classifiers more frequently for monitoring.

• English text is being prepared.Shangguan News

English translation is being prepared.

• English text is being prepared.Simon Willison

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.