Ant International’s Risk Tech Solutions Win WDTA AI STR-02 Certification

At the DWC AI STR Forum (Digital World Conference: AI Safety, Trust, and Responsibility Forum) by The World Digital Technology Academy (WDTA) in Hong Kong, AlipayHK became one of the first global tech operators to obtain a security certification on Large Language Model (LLM) applications in compliance with WDTA AI STR-02 Standards. This certification testifies to the strength of risk tech capabilities of Ant International, and marks a significant step in AI governance for Hong Kong.

This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20250305979238/en/

WDTA, a non-governmental organization operating under the UN framework, issues The LLM Security Certification as part of its AI Safety, Trust, and Responsibility (AI STR) series of standards. This certification effectively identifies and assesses issues and risks, such as data leaks, model tampering, and supplier non-compliance, throughout the entire lifecycle of large models.

Dr Tianyi Zhang, General Manager of Global Risk Management and Cybersecurity at Ant International, says: “This certification demonstrates that our ecosystem's AI capabilities are not only technologically innovative, but also integrate safety, reliability, and controllability throughout the product development process. I am glad we’ve been able to balance risk prevention, user experience, and cost in a real-world product that serves millions of users.”

By leveraging Ant International's Risk S.H.I.E.L.D (Security with High standard, Intelligent Efficiency, Lean and Dependability) strategy, AlipayHK provides 24/7 real-time risk detection and scanning, ensuring the security of user accounts and transactions. The AI-based risk control algorithm models enable dynamic risk identification and adjustment. Adaptive strategy optimization allows for flexible responses to complex and changing risk situations, providing rock-solid security protection and building a strong defense for user accounts and funds.

The large model safeguard system conducts 500,000 automated assessments daily, comprehensively identifying security vulnerabilities from dimensions such as data security, content security, and technological ethics.

• 24/7 Security Scanning: Using multimodal detection technology, it accurately identifies deepfake content such as face swaps and AIGC-generated synthetic content, and provides authenticity scores for images, audio, and text.

• Intelligent Vulnerability Management: Based on an expert vulnerability database and reachability analysis, it performs in-depth scans of large model files and provides dynamic repair recommendations.

• AI vs. AI: Utilizing large models for real-time decision-making, it blocks new types of attacks such as data leaks and model poisoning.

Ant International is a global integrated TechFin platform headquartered in Singapore, providing cross-border payment and account services for businesses, and expanding into digital growth and scenario-based financial solutions. In 2024, Ant International began operating independently and announced that it will strengthen the application and innovation of AI in the fintech sector to support the growth of emerging markets and small and medium-sized enterprises (SMEs).

Currently, Ant International's risk technology, wallet technology, payment technology, platform technology, and credit technology capabilities serve a network that includes AlipayHK and 35 other global e-wallets. These services cover more than 200 markets, reaching over 1.7 billion consumer accounts and 90 million merchants, and support more than 100 currencies.

About Ant International

Headquartered in Singapore, Ant International powers the future of global commerce with digital innovation for everyone and every business to thrive. In close collaboration with partners, we support merchants of all sizes worldwide to realize their growth aspirations through a comprehensive range of tech-driven digital payment and financial services solutions.

View source version on businesswire.com: https://www.businesswire.com/news/home/20250305979238/en/