Veriserve to expand post-launch security monitoring and improvement services for AI agents in December
Veriserve扩展智能体上线后安全监测服务
AI summary
Veriserve will launch an AI agent quality assurance solution in December 2026, extending the pre-release evaluation that began in June 2026 to post-launch monitoring, cause analysis, improvement and re-evaluation. The service is implemented in five steps: risk governance, release determination, continuous monitoring, cause analysis and improvement, and re-evaluation, and it checks security risks from perspectives including permission controls, guardrails, audit logs and prompt injection countermeasures; monitoring and one analysis are performed by AI, while impact judgment and cause analysis are handled by QA analysts.