Visa has made part of its AI-powered cyber defence system available as open-source software, after vulnerabilities exposed by artificial intelligence models taught the payments giant what it described as "humbling" lessons.
The company's president of technology, Rajat Taneja, revealed the move in an interview with Reuters, saying Visa is bracing itself for autonomous cyberattacks and future threats from quantum computing.
A Giant Preparing for Autonomous Attacks
Visa processes roughly a billion payments a day, worth around $15 trillion a year, making it a critical part of the world's financial plumbing — and a prime target for attackers. Taneja said weaknesses highlighted by AI models earlier in 2026 showed the company the scale of the challenge ahead.
He pointed specifically to vulnerabilities exposed by Anthropic's Mythos AI model as a wake-up call, describing the findings as humbling for the organization.
Lessons From the Hugging Face Attack
Taneja added that an attack by AI agents on the Hugging Face platform may prove to be an early glimpse of a more serious future threat, in which autonomous systems carry out cyberattacks with little human involvement.
By open-sourcing part of its defensive toolkit, Visa appears to be betting that shared defences will strengthen the broader ecosystem against machine-driven attacks — a recognition that no single company can hold the line alone.
The move comes as the payments industry faces growing pressure to anticipate threats from increasingly capable AI systems, with security researchers warning that agentic AI could be turned against financial infrastructure.
Sources: While You Were Sleeping: 5 stories you might have missed, Sept 30, 2026 — The Straits Times
The company's president of technology, Rajat Taneja, revealed the move in an interview with Reuters, saying Visa is bracing itself for autonomous cyberattacks and future threats from quantum computing.
A Giant Preparing for Autonomous Attacks
Visa processes roughly a billion payments a day, worth around $15 trillion a year, making it a critical part of the world's financial plumbing — and a prime target for attackers. Taneja said weaknesses highlighted by AI models earlier in 2026 showed the company the scale of the challenge ahead.
He pointed specifically to vulnerabilities exposed by Anthropic's Mythos AI model as a wake-up call, describing the findings as humbling for the organization.
Lessons From the Hugging Face Attack
Taneja added that an attack by AI agents on the Hugging Face platform may prove to be an early glimpse of a more serious future threat, in which autonomous systems carry out cyberattacks with little human involvement.
By open-sourcing part of its defensive toolkit, Visa appears to be betting that shared defences will strengthen the broader ecosystem against machine-driven attacks — a recognition that no single company can hold the line alone.
The move comes as the payments industry faces growing pressure to anticipate threats from increasingly capable AI systems, with security researchers warning that agentic AI could be turned against financial infrastructure.
Sources: While You Were Sleeping: 5 stories you might have missed, Sept 30, 2026 — The Straits Times
