ZAST.AI Raises $6M Pre-A to Scale “Zero False Constructive” AI-Powered Code Safety

The Hacker InformationFeb 10, 2026Utility Safety / Synthetic Intelligence

January 5, 2026, Seattle, USA — ZAST.AI introduced the completion of a $6 million Pre-A funding spherical. This funding got here from the well-known funding agency Hillhouse Capital, bringing ZAST.AI’s whole funding near $10 million. This marks a recognition from main capital markets of a brand new resolution: ending the period of excessive false optimistic charges in safety instruments and making each alert genuinely actionable.

In 2025, ZAST.AI found a whole lot of zero-day vulnerabilities throughout dozens of in style open-source initiatives. These findings have been submitted by authoritative vulnerability platforms like VulDB, efficiently leading to 119 CVE assignments. These should not laboratory targets, however production-grade code supporting world companies. Affected well-known initiatives embrace extensively used parts and frameworks reminiscent of Microsoft Azure SDK, Apache Struts XWork, Alibaba Nacos, Langfuse, Koa, node-formidable, and others.

It was exactly inside these extensively adopted open-source initiatives that ZAST.AI found a whole lot of actual, exploitable vulnerabilities accompanied by executable Proof-of-Idea (PoC) proof. Maintainers of those initiatives from high expertise firms like Microsoft, Apache, and Alibaba have already patched their code based mostly on the PoCs submitted by ZAST.AI.

“Within the conventional area of code safety evaluation, excessive false optimistic charges have lengthy been a core ache level plaguing enterprise safety groups. Safety engineers typically spend vital time manually verifying alerts generated by instruments, leading to extraordinarily low effectivity,” stated Geng Yang, Co-founder of ZAST.AI. “‘Report is affordable, present me the POC!’ This was the unique intention behind founding ZAST.AI — we imagine solely verified vulnerabilities are price reporting.”

ZAST.AI’s core innovation lies in its “Automated POC Era + Automated Validation” technical structure. Not like conventional static evaluation instruments, ZAST.AI leverages superior AI expertise to carry out deep code evaluation on purposes. It can’t solely mechanically generate Proof-of-Idea (PoC) code for exploiting vulnerabilities but in addition mechanically execute and confirm whether or not the PoC efficiently triggers the vulnerability. The ultimate report solely presents actual vulnerabilities which were virtually verified, attaining a breakthrough “zero false optimistic” impact.

“This is not an optimization—it is a reconstruction,” stated a consultant from Hillhouse Capital. “ZAST.AI has redefined the usual for vulnerability validation, shifting from ‘potential threat’ to ‘confirmed vulnerability, right here is the PoC.’ This adjustments the sport.”

Relating to vulnerability protection, ZAST.AI not solely helps the detection of “syntax-level” vulnerabilities reminiscent of SQL Injection, XSS, Insecure Deserialization, and SSRF but in addition possesses the aptitude to establish semantic-level vulnerabilities. This contains complicated enterprise logic flaws like IDOR, privilege escalation, and fee logic vulnerabilities—areas lengthy thought of tough for automated instruments to achieve. Think about your safety software crying “wolf” day-after-day, with a false optimistic fee above 60%. By the point the actual “wolf” seems, the group may already be desensitized. This is not a folks drawback; it is a software defect—they’ll solely speculate, not show.

Presently, ZAST.AI already serves a number of enterprise purchasers, together with Fortune International 500 firms. By mechanically discovering unknown vulnerabilities and instantly offering runnable PoC vulnerability experiences, ZAST.AI helps purchasers considerably shorten vulnerability remediation cycles, markedly scale back safety operation prices, and has gained excessive recognition from clients. This spherical of funding will primarily be used for core expertise R&D, product function enlargement, and world market improvement. CEO, Geng Yang said: “Our imaginative and prescient is to construct an end-to-end AI-driven safety platform, enabling each improvement group to acquire the best high quality safety assurance on the lowest value. Sooner or later, ZAST.AI will proceed to deepen technological innovation in AI + Safety, offering world clients with smarter, extra exact, and extra environment friendly code safety options.”