Hong Kong FinTech Week: HKU Releases the World's First KYA Report, with Corporate Practices from Ant Digital Technologies, GCSA and Others Becoming Industry Benchmarks During the recent Hong Kong FinTech Week × StartmeupHK Festival 2026, Professor Yang Wang, Vice-President of the University of Hong Kong (HKU), and Associate Professor Hong Yang of HKU Business School, representing the HKU KYA Research Group, officially released the KYA: Know Your Agent — Research Report on the Agent Credit System. The Global Cybersecurity Alliance (GCSA) was featured as a representative industry case in cybersecurity protection and offensive-defensive security governance. Under the central proposition that “Without KYA, there can be no agent economy,” the report highlights how the critical challenge facing the emerging agent economy is shifting from technological capability to trust as artificial intelligence evolves from content generation to autonomous action. It introduces a three-dimensional framework encompassing trusted identity (Identity), trusted credit (Credit), and trusted permissions (Permission), alongside an L0–L5 agent maturity model, providing a systematic foundation for agent credit assessment and trustworthy governance. According to the report, 76% of survey respondents had encountered concerns related to agent trust, security, or compliance over the preceding year, while 88% believed the industry needs unified standards for evaluating or certifying the trustworthiness of AI agents. In Chapter Four, "Practical Validation," the report features a dedicated case study titled "GCSA: Security Protection and Offensive-Defensive Governance for the Agent Ecosystem." The study examines GCSA's technical practices in AI code security, data loss prevention, agent-based vulnerability auditing, and automated offensive security evaluation. Among the initiatives highlighted is SecVibe Code, jointly developed by the AI & Cybersecurity Research Institute established by HKU and GCSA. Built on a Harness architecture, SecVibe Code integrates dynamic security auditing and data loss prevention (DLP) capabilities directly into AI-assisted coding workflows, helping developers identify code vulnerabilities, sensitive credentials, and potential privacy risks. The report also references CyberGym benchmark results provided by GCSA, highlighting the research progress of its security auditing agents in vulnerability identification, attack-chain analysis, and validation. GCSA emphasizes in the report that security data constitutes an essential component of an agent credit system. However, building a comprehensive KYA framework is beyond the capabilities of any single organization. It requires collaboration among governments, research institutions, and industry stakeholders to advance common standards, data sharing, and coordinated governance. By featuring practices from Ant Digital Technologies, Stripe, GCSA, and other organizations across different sectors, the report illustrates the diverse approaches emerging in the development of a trustworthy agent ecosystem. Looking ahead, GCSA will continue advancing AI security research and offensive-defensive security technologies, while actively contributing to the development of trustworthy agent governance frameworks to support the secure and sustainable growth of the agent economy. 📖 Full Report: Stay Connected with GCSA Follow GCSA on our official social media channels for the latest news, research updates, and developments. ✈️ Telegram Channel: 🫧WeChat Official Account: #GCSAInsights# #GCSAcybersecurity# #HKUweb3#