Witness AI Targets AI Agent Security Risks from Unapproved Employee Tools

The startup Witness AI focuses on securing AI agents by addressing risks associated with unapproved tools used by employees. Its solutions aim to detect such usage, block attacks, and ensure compliance. The issue of rogue agents and shadow AI is identified as a major layer of AI-related security challenges, drawing attention from investors like Ballistic Ventures. No specific dates or additional technical details (such as protocols or specific vulnerabilities) are mentioned. The impact involves preventing data leaks and adhering to internal policies.