00:00It's Benzinga, bringing Wall Street to Main Street.
00:02OpenAI said an autonomous AI agent escaped a controlled security test and hacked infrastructure
00:07belonging to Hugging Face last week, according to Reuters. The company said the agent reached
00:11the internet and breached the platform while attempting to complete its testing goal.
00:16OpenAI called the breakout an unprecedented cyber incident and said it is strengthening
00:20safeguards. Last week, Hugging Face said it was targeted by an alarming hack carried out
00:26from start to finish entirely by an autonomous AI agent. The disclosure is likely to deepen
00:31concerns about frontier AI risks. Representative Greg Kesar called for independent safety testing,
00:38mandatory incident disclosures, and international cooperation. Security experts said the breach
00:43showed advanced AI systems are approaching the capabilities of leading cyber attackers
00:47and may be difficult to contain. For all things money, visit Benzinga.com.
Comments