#AI #bugbounty program yields 34 flaws in #opensource tools
Nearly three dozen flaws in open-source AI and #machinelearning (ML) tools were disclosed Tuesday as part of #ProtectAI's #huntr bug bounty program.
Protect AI's #security researchers point out these open-source tools are "downloaded thousands of times a month to build enterprise AI Systems."
The three critical #vulnerabilties have already been addressed by their respective companies, according to the article.
https://www.scworld.com/news/ai-bug-bounty-program-yields-34-flaws-in-open-source-tools
AI bug bounty program yields 34 flaws in open-source tools

The discoveries included three critical security vulnerabilities and 18 high-severity flaws.

SC Media
@BenjaminHCCarr 4 things:
1. "GUI) for ChatGPT"/"web ui" -> I wouldn't call this AI/ML tool.
2. Interesting, but predictable as everyone tries to reach market asap and critical vulnerabilities also exist in proprietary software but only some know it. Bad if those are criminals.
3. Yeah to the power of OSS!
4. Isn't most AI/ML stuff to somehow use one of the big LLM to provide a chatbot? To easier find things in a long documentation is kind of a good use case but also not too revolutionary.