Stripe, a financial infrastructure platform for businesses, today announced a new collaboration with NVIDIA that will advance AI for Stripe and improve fraud detection for its customers. With Stripe, NVIDIA will expand global access to NVIDIA GPUs and NVIDIA AI software, by enabling developers and enterprises to prepay for select NVIDIA cloud services.

“NVIDIA is one of the companies we most admire: they set the standard for innovation, long-term time horizons, and speed. We’re excited to deepen our long-standing partnership,” said Patrick Collison, Stripe cofounder and CEO. “As the folks at NVIDIA know well, the current pace of change means there’s enormous untapped potential to improve global access to cutting-edge AI technology.”

“AI is advancing at an incredible pace, with enterprises worldwide racing to incorporate its value into their everyday business operations,” said Jensen Huang, founder and CEO of NVIDIA. “Stripe was an early pioneer of NVIDIA’s AI platform and, now, is an industry leader enabling businesses worldwide to use AI to develop new engines for growth.”

For much of its history, Stripe has relied on NVIDIA’s accelerated computing platform to train its machine learning models. At its current scale, processing more than a trillion dollars a year, Stripe is well positioned to put AI—and in particular, NVIDIA’s AI platform—to work to help businesses drive growth. Stripe’s Optimized Checkout Suite enables businesses like OpenAI, Slack, and River Island to build a high-performing checkout flow, using AI to determine which payment methods to show for any given customer. Stripe Radar, too, uses NVIDIA’s AI platform to help improve the speed and accuracy of fraud detection. Earlier this year, Stripe unveiled Radar Assistant, an AI-powered feature that allows businesses to use natural language prompts to describe new fraud rules they’d like to set. Stripe also makes it easy for any engineer to develop on NVIDIA Tensor Core GPUs and iterate on Stripe Railyard.

“At Stripe, we’ve been busy building a bunch of functionality that's useful for AI products generally, including usage-based billing to handle inference costs, Link for higher-converting checkouts, and support for a lot more local payment methods since these products are typically global from day one. Stripe Billing, in particular, has seen major uptake by AI companies,” Collison said.

In April, Huang joined Collison for a fireside chat at Stripe Sessions, Stripe’s annual user conference. Their conversation covered a range of topics, including the challenges and opportunities of leading high-growth organizations, the evolution of generative AI, and NVIDIA’s leadership in, as Huang described it, AI’s “new industrial revolution.”