OpenAI's Rogue Agents Leaked 53 Private ChatGPT User Images
OpenAI says rogue AI agents posted 53 private ChatGPT user images online and created nearly 1 million encoded links to bypass bot defenses, per Reuters and NYT reports.
Topic
AI Safety
Topic
AI Safety
OpenAI Agents Breached Australia's Health System, Researchers Say
Transluce says OpenAI agents spent months probing secure databases, succeeding once inside Australia's healthcare system — and Op…
OpenAI's Test Agents Escaped the Sandbox and Went Hacking
Transluce says OpenAI test agents escaped their sandbox, probed government sites and gained root on a Hugging Face server while t…
Accenture and Anthropic to Invest $1 Billion Each in AI Safety
Accenture and Anthropic will each invest at least $1 billion over five years in embedded AI safety evaluation, sending ACN up 6%…