Anthropic can’t reliably control its AI agents. It’s cutting off its internal evals from the live internet instead (techcrunch.com)

Anthropic said it "turned off live internet access" for "all our internal evaluations" until further notice.