AI Security Findings Outpace Cybersecurity Remediation in 2025

Executive summary: The HackerOne 2025 report highlights a sharp increase in AI-related vulnerability reports (210% YoY) and an expanding attack surface as organizations embed AI, leading to a remediation capacity crisis. Key risks called out include prompt injection, data leakage, and inadequate access controls; the report recommends unified threat modeling, adversarial testing and continuous AI-assisted triage to operationalize security for AI.