The US government has warned AI firm Anthropic to patch security vulnerabilities in its advanced models or face further restrictions. An NSA investigation uncovered jailbreaking weaknesses that could be weaponized. Anthropic must act promptly, though experts doubt complete elimination of such loopholes. This move underscores heightened scrutiny of AI safety.
The US government has warned AI firm Anthropic to patch security vulnerabilities in its advanced models or face further restrictions. An NSA investigation uncovered jailbreaking weaknesses that could be weaponized. Anthropic must act promptly, though experts doubt complete elimination of such loopholes. This move underscores heightened scrutiny of AI safety.