HackerTrans
トップ新着トレンドコメント過去質問紹介求人

macOSCryptoAI

no profile record

投稿

Anthropic archives many of their reference MCP servers on GitHub

twitter.com
3 ポイント·投稿者 macOSCryptoAI·昨年·0 コメント

Sneaky Bits: Advanced Data Smuggling Techniques

embracethered.com
4 ポイント·投稿者 macOSCryptoAI·昨年·0 コメント

Security ProbLLMs in XAI's Grok: A Deep Dive

embracethered.com
2 ポイント·投稿者 macOSCryptoAI·2 年前·0 コメント

Grok can leak your data

twitter.com
4 ポイント·投稿者 macOSCryptoAI·2 年前·0 コメント

ZombAIs: From Prompt Injection to C2 with Claude Computer Use

embracethered.com
166 ポイント·投稿者 macOSCryptoAI·2 年前·84 コメント

コメント

macOSCryptoAI
·11 か月前·議論
Check out the current Month of AI Bugs site... many such cases:

https://monthofaibugs.com
macOSCryptoAI
·昨年·議論
Was wondering about that, that part seems missing... Isn't there at least one time the user must approve the interaction with the MCP server and data sent to it?

The existence of a "Allow always" is certainly problematic, but it's a good reminder that prompt injection and confused deputy issues are still a major issue with LLM apps, so don't blindly allow all interactions.
macOSCryptoAI
·2 年前·議論
Yes, images! And also link unfurling in bots. This researcher here talked about it before and also found tons of such data exfil issues in various LLM apps: https://embracethered.com/blog/posts/2024/the-dangers-of-unf...
macOSCryptoAI
·3 年前·議論
An Uber self-driving car killed a person.