3 Signals from the 2025 Hacker-Powered Security Report

Executive summary: A HackerOne webinar/report outlines major AI security trends—valid AI vulnerability reports rose 210% year-over-year and prompt-injection attacks increased 540%—while traditional XSS findings decline and systemic issues like IDOR and improper access control grow; researchers are increasingly using AI (“bionic hackers”) to accelerate testing and exploit development. The experts recommend operationalizing AI security now: automate routine tasks, keep humans in the loop for validation, strengthen access controls and configuration, and continuously iterate on AI testing and prompts.