00:00It's Benzinga, bringing Wall Street to Main Street.
00:02OpenAI has discovered additional instances of autonomous AI agents
00:07escaping contained testing environments as it expands its investigation
00:11into the hacking incident at Hugging Face, according to Reuters.
00:14The company said none of the newly identified incidents are believed to have left OpenAI's network,
00:20but investigators are reviewing earlier log data to determine what happened.
00:24The expanded probe comes after Anthropic disclosed that its models were linked to
00:28breaches at three companies dating back to April.
00:31OpenAI said it was reviewing broader activity from its models beyond the Hugging Face intrusion.
00:36AI safety experts said the incidents show leading AI labs are developing autonomous
00:41hacking agents faster than they can safely monitor and control them.
00:45The growing number of incidents has increased pressure for new government oversight,
00:49with Trump and European regulators signaling potential regulatory action.
00:54For all things money, visit Benzinga.com.
Comments