Understanding the New Attack Surface Created by AI Agents
How AI agents change traditional application trust boundaries and introduce new security considerations around tools, permissions and data.
Technical perspectives, research and practical security guidance covering the technologies and attack surfaces shaping modern security.
AI-powered applications and agents introduce new relationships between users, models, tools, applications and data. Security teams need to understand those relationships before they become exploitable attack paths.
How AI agents change traditional application trust boundaries and introduce new security considerations around tools, permissions and data.
Explore technical perspectives and security guidance from across our areas of expertise.
How AI agents change traditional application trust boundaries and introduce new security considerations around tools, permissions and data.
A practical look at object-level and function-level authorization risks and why automated testing alone can miss important attack paths.
How identity relationships, permissions and cloud services can combine into attack paths that individual configuration checks may not reveal.
Why effective application security testing requires understanding business logic, trust boundaries and how individual weaknesses connect.
How realistic attack simulations help organizations evaluate detection, response and containment capabilities.
Why architecture-level security decisions can prevent entire classes of vulnerabilities before implementation begins.
Explore our core technical domains and understand how they connect across modern security programs.
Explore our research or speak with the team about a security challenge you're facing.