Ask HN: How do you handle AI security for automated vulnerability management?

A cybersecurity newcomer used an AI-powered tool for automated vulnerability detection and patching, but encountered an issue when the tool proposed a fix for a critical issue that was later disputed by another entity, Gemini. This disagreement led to a loss of trust in the platform. The experience suggests that AI security platforms for repositories may be unreliable. The user now believes that cross-checking every proposed fix is necessary. This incident highlights a potential common issue with AI security platforms that provide automated vulnerability management. The user's experience has led them to question the reliability of these platforms.