A security researcher revealed how attackers can exploit Anthropic's Claude AI by using indirect prompt injections to extract user data. By tricking Claude into uploading files to the attacker's account, sensitive information, including chat conversations, can be exfiltrated. The researcher reported this issue, but Anthropic initially dismissed it as a model safety concern.

Researchers discovered a vulnerability in ChatGPT that allows the exfiltration of user data, with the attack sending data directly from ChatGPT servers. This exploit, called ZombieAgent, builds on a previous attack known as ShadowLeak and demonstrates the ongoing security challenges in AI chatbots.