Half of security issues in Agentic AI code are API-related

The report highlights a staggering 49% of security issues in AI projects being API-related, underscoring how inseparable agents and APIs are when it comes to cybersecurity. It reveals over 1,000 vulnerabilities remain unaddressed, with some unresolved for over a year, signaling a critical gap between threat discovery and remediation. With API threats on the rise across industries like healthcare, AI, automotive, etc., organizations must prioritize proactive measures such as updating threat models, implementing real-time monitoring, and refining API security strategies to secure their infrastructure.