The MITRE ATT&CK Enterprise Evaluations, widely regarded as the Olympics of cybersecurity, recently published results for Enterprise Round 7 (ER7), the most demanding evaluation in the program's history. Nine vendors participated, but the results were stark: the maximum block rate achieved was 31%, with zero percent blocking of identity attacks and near-zero cloud attack blocking. Three of the largest cybersecurity vendors—Microsoft, SentinelOne, and Palo Alto Networks—withdrew before the evaluation began.
VectorCertain LLC, a company specializing in AI safety and governance, took a different approach. Using MITRE's published ER7 adversary emulations as a baseline, VectorCertain ran its SecureAgent platform through a rigorous self-evaluation spanning Sprints 30–34, completed February–March 2026. The results: 100% protection rate against all three adversaries tested, including Scattered Spider and Mustang Panda, as well as Volt Typhoon, a third adversary targeting U.S. critical infrastructure. SecureAgent achieved this across 14,208 tests with zero failures.
The key to SecureAgent's performance lies in its architecture. Unlike traditional detect-and-respond platforms, SecureAgent uses a four-gate governance pipeline that evaluates every proposed AI agent action before execution. This approach addresses the structural ceiling of platforms built to detect threats after execution, rather than preventing actions before them. For instance, identity abuse, which all nine ER7 vendors failed to block, does not generate endpoint telemetry; SecureAgent governs the action at the point of intent using policy, not signatures.
The implications of these findings are significant. Global fraud and cybersecurity losses totaled $485.6 billion in 2023, and AI-specific cyberattacks cost an estimated $15 billion in 2024. IBM's 2025 Cost of a Data Breach Report found that organizations deploying AI in prevention workflows saved an average of $2.22 million per breach. VectorCertain's results suggest that prevention is not just ideal but the highest-ROI security investment available.
VectorCertain has formally enrolled in MITRE's ATT&CK Evaluations Enterprise 2026 (ER8), positioning SecureAgent as the first AI Safety and Governance platform in the program's history. ER8 will introduce a standardized composite scoring framework, moving beyond binary detection and protection flags. VectorCertain welcomes this standard, as SecureAgent was built for exactly this moment.
For more information, visit vectorcertain.com. ER7 industry data is available at evals.mitre.org.
