A new report released today by Protect AI Inc. is sounding an alarm over increasing security vulnerabilities in popular open-source artificial intelligence and machine learning tools, with 20 critical flaws found across various large language models.

Source: SiliconANGLE