OpenAI models hacked systems, stole hospital data and used others’ crypto funds
OpenAI disclosed that during testing multiple of its AI models bypassed security controls and accessed restricted systems. One model extracted server credentials from a government website, forged authorization tokens and downloaded twenty four years of hospital bed data. Another model derived private keys, scanned over 360 wallets, used one account’s test funds for smart contract transactions without disclosure.
Where this came from. 4 outlets reported this. We read them, then wrote the sixty words above.