Security Professionals Cite AI as Top Security Risk

A HackerOne survey of 500 security professionals reports that 48% view AI as the top security risk, citing training data leakage (35%), unauthorized AI use (33%), and external compromise of AI models (32%), with 58% expecting an AI-driven arms race between attackers and defenders. Experts from HackerOne, Bugcrowd, and Zendata stress AI’s dual-use nature—enabling automated attacks and adaptive malware while also accelerating defense—and recommend stronger governance beyond RBAC, data fingerprinting, rigorous testing to prevent data leakage, and greater internal deployment of open-source and domain-specific models.