Conversation

“New hack uses prompt injection to corrupt Gemini’s long-term memory”

https://arstechnica.com/security/2025/02/new-hack-uses-prompt-injection-to-corrupt-geminis-long-term-memory/?utm_source=bsky&utm_medium=social

This is why I don't bother commenting on "AI" news much anymore. People have been pointing out that this shit is badly thought out and inherently insecure since before the ChatGPT launch, and the news keeps confirming it, but none of it registers with the bubble crowd.