A recent Infosecurity Magazine report highlights a growing divide between the rapid adoption of agentic AI by malicious actors and the cautious stance of enterprise defenders. While some organizations utilize AI to streamline threat detection and log analysis, security teams remain wary of giving automated agents direct control over critical response and remediation actions. Industry experts point to persistent fears regarding hallucinations, prompt injection vulnerabilities, and a lack of reliable governance frameworks for autonomous non-human identities. Conversely, threat actors are successfully deploying agentic workflows to orchestrate automated vulnerability scans, exploit targets, and drastically lower the operational costs of cybercrime. Cybersecurity leaders emphasize that building trust in agentic defense tools will require strict guardrails, transparent decision chains, and continuous human-in-the-loop oversight.

Full report : New findings reveal that while cybersecurity professionals increasingly leverage AI for reconnaissance, many remain hesitant to grant autonomous agents high-level decision-making authority over security operations.