Many LLMs use teaser-phrasing to get users to keep going in a conversation. OpenAI says they are reducing this in ChatGPT.
Infosecurity spoke to several experts to explore what CISOs should do to contain the viral AI agent tool’s security vulnerabilities ...