A new cheat sheet from the REMnux project outlines how malware analysts can direct an AI agent through a contained, network-isolated lab and verify its findings, using tools such as the REMnux MCP ser…
A survey of over 300 security professionals, conducted by Sounil Yu and co-author Adrian Sanabria and published as the AI Security Decisions Report, found that organizations most often secure AI with …
Only 41% of organizations reported using generative AI for security-related tasks under strict policy, according to the 2026 SANS AI Survey, prompting security leaders to adopt the Security Autonomy M…
Sounil Yu and the AI Defense Matrix Catalog identify five questions buyers should answer before purchasing AI security products, including checking which asset the product protects, determining whethe…
A new AI Defense Matrix Catalog maps over 100 AI security products into four clusters, revealing that runtime guardrails dominate while Respond and Recover capabilities are scarce. Sounil Yu and the a…
Cyber Company Profiles, a new site created by cybersecurity expert Lenny Zeltser, scores hundreds of cybersecurity companies based on public sources using standardized frameworks. The site offers free…
The AI Defense Matrix Catalog, maintained by Sounil Yu and an unnamed co-author, lists 239 products defending AI systems, revealing a lopsided market where most products cluster around runtime AI data…
Sounil Yu and an unnamed co-author launched a five-minute, ten-question survey to benchmark how organizations secure AI, covering which AI assets are protected, who decides, and where controls come fr…