Agentic AI Sparks Urgent Security Concerns Across Industries

Agentic AI, capable of autonomous reasoning and decision-making, is driving a technological revolution but it also introduces significant security risks. As outlined in the 2025 Identity Security Landscape report, 68% of organizations lack robust identity controls for AI systems, leaving critical vulnerabilities exposed. These agents are increasingly deployed in cybersecurity operations (e.g., AI SOC analysts) to fight threats, but they can themselves be subverted via prompt injections, lateral movement, and malicious task chaining.

Cyber experts warn that attackers may exploit agentic AI’s autonomy to scale phishing campaigns or launch automated exploits. Defense will require advanced measures: layered access controls, behavioral monitoring, input sanitization, and new threat models tailored for autonomous agents. As AI agents proliferate across industries, the race is on to build effective frameworks, combining technical safeguards, identity governance, and ethical oversight to secure both the promise and peril of agentic intelligence.

Source: indiatimes.com