C2A Security's New EVSec Release Leverages AI to Put the Threat Model Behind Every Vulnerability Decision
News originally published by
EIN PresswireNew vulnerabilities are judged against a threat model built from the team's own method, with the evidence shown before a person decides
AI does the analysis, the evidence is visible, and the decision stays with the people responsible for product security.”
JERUSALEM, ISRAEL, November 17, 2026 /EINPresswire.com/ -- JERUSALEM, Israel, November 17, 2026 – C2A Security, provider of EVSec, the AI-driven, context-aware product security orchestration platform for software-defined products and cyber-physical systems, today announced a new EVSec release that uses AI to connect threat modeling, vulnerability prioritization, and human decision-making in one continuous chain. — John Heldreth
The release builds the threat model from each team's own method and the product it actually ships. It assesses new vulnerabilities against that model and shows the evidence behind each recommendation, so a person makes the final call and the decision stays on record. For product security teams, the release removes the work they least want to do. Instead of rebuilding each ECU, device, or system by hand in yet another tool, teams can reuse their own methods across projects, keep threat models current as development moves forward, and produce consistent results regardless of who builds them.
UN R155 requires vehicle manufacturers to keep cybersecurity risk assessments current. ISO/SAE 21434 defines how that risk is managed across the vehicle lifecycle. FDA Section 524B requires manufacturers of covered cyber devices to submit plans for monitoring and addressing postmarket vulnerabilities. The EU Cyber Resilience Act already requires reporting of certain actively exploited vulnerabilities and security incidents, with broader vulnerability-handling obligations taking effect in December 2027. IEC 62443-4-1 sets secure development lifecycle requirements for industrial components.
Across these frameworks, manufacturers need to demonstrate which risks affect their products, what actions they have taken, and why. In practice, most teams still build threat models by hand over weeks. They then judge incoming vulnerabilities by a severity score that knows nothing about the product they are sitting in.
Tools that build the threat model and tools that rank vulnerabilities usually sit apart, leaving teams to connect them by hand and justify the result to assessors and customers.
"Product security teams are not short of findings. They are short of time," said John Heldreth, VP Product at C2A Security. "They spend weeks building a threat model by hand, hours on vulnerabilities that have no impact on their product, and days assembling evidence when an assessor asks for it. This release connects those steps, so a team can go from a new finding to its actual impact on their product without the manual handoffs. AI does the analysis, the evidence is visible, and the decision stays with the people responsible for product security."
EVSec builds a contextual cyber model of each product, connecting architecture, components, vulnerabilities, threats, compliance obligations, and risk workflows. The new release brings three steps together in one continuous workflow, from design to decision:
• A threat model built from your method: AI-guided workflows turn a team's answers and its product context into a working threat model, from systems and assets to attack trees. The method is captured once and reused on every project, so the result is consistent no matter who builds it. Threat analysis that used to take weeks takes hours.
• Priorities set by your product, not a generic score: AI assesses new SBOMs, vulnerabilities, and threat intelligence against the model, so the same vulnerability can be critical in one product and irrelevant in another, and EVSec shows which, and why. Ruled-out findings keep their reasoning and are revisited when the SBOM or threat picture changes. EVSec assesses the impact of a new vulnerability in seconds. In customer deployments, it has reduced vulnerability noise by 98 to 99 percent.
• Every recommendation shows its work, and a person decides: Each recommendation carries its evidence: the vulnerability, the attack step it changed, the resulting risk, and the reasoning and confidence behind it. EVSec's AI runs the analysis; a designated reviewer decides how the risk is treated, and that
decision and its rationale are recorded. The assessment, decisions, and evidence export as a single package for audits and regulatory documentation.
For teams whose code cannot leave their own environment, the release also introduces the EVSec Edge Analyzer. It generates SBOMs from binaries and source code inside the customer's own environment, without transferring proprietary code or binaries to the EVSec cloud. The resulting component information feeds EVSec's threat modeling and vulnerability prioritization.
What sets the release apart is that each step builds on the one before: every vulnerability is ranked against a model of the product the manufacturer actually ships, not a template or a declared list of components.
The new EVSec release is available now to EVSec customers in automotive, medical device, and industrial markets. To learn more or schedule a demonstration, visit c2a-sec.com or contact us at contact@c2a-sec.com.
About C2A Security
C2A Security provides AI-driven, context-aware product security orchestration for software-defined products in heavily regulated industries, including healthcare, industrial, robotics, automotive and other cyber-physical systems. Its customers and partners include Bayer, BMW Group, Daimler Truck AG, Ascensia, Elekta, NVIDIA, Siemens, Orcanos, HARMAN, Marelli, NTT Data, and Deloitte. C2A Security is a CLEPA Innovation Award recipient for its DevSecOps platform. Founded in 2016 by NDS/Cisco veteran Michael Dick, the company is headquartered in Jerusalem, Israel. c2a-sec.com
David Leichner
C2A Security
Visit us on social media:
LinkedIn
Legal Disclaimer:
EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
