Visa Uses AI to Cut Some Cybersecurity Fixes From Weeks to Hours
Visa has enhanced its AI cybersecurity framework, VVAH, to accelerate vulnerability remediation and validation. The upgraded system can now resolve some attack paths in hours, a significant reduction from the previous weeks-long process. This open-source framework, initially developed after participating in Anthropic’s Project Glasswing, supports models from Anthropic, OpenAI, and others. Visa also introduced three new advisory services covering AI cybersecurity education, VVAH-based maturity assessments, and cyber risk roadmaps. Tens of thousands of developers have downloaded VVAH since its open-source release in June 2026.
Visa’s VVAH upgrade points to a critical shift in cybersecurity strategy across Asia. As AI-enabled threats accelerate, the ability to move from vulnerability discovery to validated remediation in hours, rather than weeks, becomes a competitive necessity. This speed is particularly relevant for financial institutions and critical infrastructure providers in markets like Singapore and Hong Kong, where data security and operational continuity are paramount. The framework’s open-source nature and support for multiple AI models also suggest a collaborative approach to defense, which could foster broader adoption and innovation within the region’s tech community. The introduction of cybersecurity advisory services by Visa further underscores the growing demand for expert guidance in navigating the AI threat landscape. For Asian enterprises, these services offer a structured path to assess their cyber resilience and implement AI-powered defenses. The collaboration with NVIDIA’s Open Secure AI Alliance and IBM/Red Hat’s Project Lightwell also indicates a move towards integrating VVAH into a wider ecosystem of security tools, potentially enhancing its utility and reach across various industries in Asia.
