October 13, 2026: TrendAI™, the global AI security leader and enterprise business unit of Trend Micro Incorporated, announced the private preview of TrendAI Vision One™ Autonomous Vulnerability Discovery, a service powered by TrendAI’s agentic exploit-remediation engine, code name AESIR, which identifies, validates, and verifies fixes for vulnerabilities, including zero-days, in customer source code.
TrendAI’s agentic system is a purpose-built, multi-model AI architecture using primarily Anthropic models. It combines autonomous vulnerability discovery, exploit identification, and threat hunting with research from the TrendAI™ Zero Day Initiative™ (ZDI), the world's largest vendor-agnostic bug bounty program. When evaluated against the CyberGym testing framework, the engine outperformed single-model and agentic system scores on the CyberGym Leaderboard.
The service is deployed and operated on Amazon Bedrock from Amazon Web Services (AWS). Customers keep using the code repositories and version control systems they already rely on.
"AI is changing both sides of vulnerability discovery. Code is being written faster than ever, and attackers are using AI to find flaws faster than ever," said Sharda Tickoo, Country Manager, India and SAARC. "Security teams shouldn't have to choose between the speed of AI and the judgment of the world's best vulnerability researchers. Autonomous Vulnerability Discovery puts AI built for accuracy together with two decades of ZDI research, and runs it where customers’ code already lives, so they can find and fix what matters before attackers do.”
Three differentiators set the service apart.
- Detection accuracy
Internal benchmarking shows AI models from TrendAI™ identify more true-positive vulnerabilities, with fewer false positives, than leading vulnerability scanning and static analysis tools.
- Validation backed by TrendAI™ ZDI
Findings are validated by ZDI experts and checked against more than 20 years of research from the program and the team behind the renowned Pwn2Own™ hacking competition.
- Native deployment in Amazon Bedrock
The service runs on Amazon Bedrock, allowing customers to keep their existing Git repositories and code registries as the system of record, and source code stays within the customer’s AWS environment.
How it works
TrendAI™ begins with a scoping interview to understand the customer’s environment. A TrendAI™ ZDI expert uses that input to define and tune a custom scan. To ensure safety, the service team monitors and controls every scan, rather than handing customers an unsupervised black box. Once the scan runs against a customer’s source code, TrendAI™ produces a comprehensive vulnerability report that includes a severity rating, an estimated CVSS score, and a detailed write-up for every finding. Each discovery is cross checked against TrendAI™ ZDI vulnerability research. After the customer builds and deploys a patch, the scan is run again to verify the fix, once again checking it against TrendAI™ ZDI research before issuing a final verified patch status.
Because the service is deployed and operated on Amazon Bedrock, the entire workflow, from initial scan through final patch verification, can run inside the customer’s AWS environment.
Availability
Autonomous Vulnerability Discovery is available now in private preview to select enterprise customers running on AWS. Pricing is customized based on scan scope and deployment model. Customers can request a scoping interview by contacting their TrendAI™ account team.
