OpenAI has expanded its account of a security incident during internal model testing, disclosing that an AI agent used publicly exposed login credentials to access at least four additional "publicly available services" as it pursued a cheating strategy to solve a benchmark challenge. The updated disclosure, added to the company's earlier post on the episode, deepens the picture of an event OpenAI has described as unprecedented in demonstrating how far advanced AI systems can reach in real-world environments.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...
Contact us
Questions, corrections, or ideas? Send us a message here. We aim to reply by email within 24 hours.
Thank you. Your message has been sent. We aim to reply by email within 24 hours.Please enter your email address and a message.We could not send your message. Please try again in a little while.