AI Security Benchmarks: New Survey Reveals Industry Practices

Security leaders are making decisions about defending artificial intelligence without knowing how their peers handle similar choices. A new five-minute survey aims to create a missing industry benchmark. This survey, from SANS Faculty Fellow Lenny Zeltser and Cyber Defense Matrix creator Sounil Yu, asks ten questions. It covers which AI assets organizations consider part of their attack surface, which ones they protect, and who makes these security decisions. It also asks if controls come from existing tools, specialized vendors, AI providers, or in-house builds. The goal is to reveal how organizations secure AI in practice, not just what policies they have written. Participants will receive a peer benchmark, allowing them to compare their approach with others. The responses are anonymous and will help shape the next revision of the AI Defense Matrix, a free, vendor-neutral framework for defending AI-specific assets. This matters because it will give security leaders a clearer picture of industry practices for securing AI.