Summary
A new report by Reco indicates that 80% of AI tools in organizations operate without IT oversight, posing significant cybersecurity risks and potential for data exfiltration. The study found 637 vulnerabilities across AI agents and LLMs, highlighting a widespread operational risk.