Checkmarx Wins Top Award for AI-Driven Code Security
The platform's new AI-reasoning engine outperforms traditional static security benchmarks, marking a shift for automated code auditing.
Updated on Oct. 27, 2026 in Cybersecurity
Checkmarx has secured the 2026 Editor's Choice GenAI Agentic Application Security award from Cyber Defense Magazine. The award recognizes the company's platform, Checkmarx One, for its capacity to protect software developed by both human engineers and AI agents.
Why it matters
As development workflows increasingly integrate AI agents, traditional security tools struggle to keep pace with non-human code generation. This award highlights the shift toward AI-native scanning techniques designed to identify vulnerabilities in complex, automated supply chains.
Checkmarx's Fusion SAST engine achieved an F1 score of 0.74, significantly outperforming the 0.20 average recorded for traditional query-based static application security testing tools. The platform is designed to handle the scale of trillions of lines of code annually.
The players
Checkmarx
A cybersecurity firm that provides application security testing platforms and hybrid scanning engines for code and supply chain management.
Cyber Defense Magazine
An industry publication that hosts annual security awards, currently in their 14th year.
The details
Checkmarx One utilizes hybrid scanning engines that aggregate security data across both proprietary code and third-party supply chain components. Its Fusion SAST — a method of automated security analysis — incorporates frontier AI model reasoning to evaluate code logic. By moving beyond simple query-based pattern matching, the system aims to improve accuracy in detecting complex vulnerabilities in codebases generated by AI agents.
Timeline
October 27, 2026: Checkmarx announced the receipt of the award.
The Tech Race
This award marks a shift in the 14th annual Cyber Defense Awards toward prioritizing AI-native security architectures. The recognition highlights the competitive race to move past traditional query-based testing toward model-reasoning frameworks capable of auditing autonomous AI coding agents.
Developers and security teams currently using the Checkmarx One platform can expect improved detection accuracy for agent-generated code. Organizations relying on standard static testing may consider these performance benchmarks when evaluating the transition to AI-augmented security tools.
The takeaway
The performance gap between legacy scanning and AI-reasoning tools suggests that organizations should prioritize evaluating F1 scores for their own codebases. Watch for the next round of industry benchmark releases to see if these F1 gains hold across broader, non-synthetic production environments.
Further reading
Explore the Cybersecurity section for more on how automated scanning is evolving to handle AI-generated code.
Source note: This article includes information reported by The Manila times.
