Stripe co-founder John Collison on AI powering a 'good old-fashioned tech race'

Stripe co-founder John Collison discusses the company’s development of an AI foundation model trained on Stripe’s payment data to improve fraud prevention and payment success, emphasizing their focus on security and global expansion. He highlights the rapid growth of AI-driven startups, the evolving tech race among industry giants, and Stripe’s long-term, innovation-driven strategy inspired by Warren Buffett’s principles.

In the video, Stripe co-founder and president John Collison discusses the company’s recent product event, Stripe Sessions, held in San Francisco, where they unveiled their first AI foundation model for payments. This proprietary model is trained on Stripe’s extensive payment data and aims to enhance various functions such as fraud prevention and conversion optimization. Collison explains that this integrated approach allows Stripe to provide merchants with advanced inference capabilities, reducing fraud and improving payment success rates, all built from their own AI infrastructure rather than relying on external sources like OpenAI.

Collison emphasizes the evolving landscape of fraud in the age of AI, comparing it to historical challenges like automobile-related crimes. He notes that AI introduces new adversarial threats, such as deepfakes and sophisticated scams, but Stripe’s focus remains on enabling businesses to concentrate on product development while Stripe handles security and fraud prevention out of the box. The company’s goal is to simplify payment processes for developers and merchants, providing tools that work seamlessly across global markets, including stable coin payments and cross-border payouts, which are gaining traction in real-world business applications.

The discussion shifts to the rapid growth of AI companies and their impact on Stripe’s business. Collison highlights that AI-driven startups are among Stripe’s fastest-growing customers, with some reaching hundreds of millions in annual recurring revenue within a short period. He also talks about new product initiatives, such as tools for agentic buying—where AI can autonomously make purchases within applications—pointing to a future where commerce takes place directly within AI environments. Collison believes this shift will change consumer behavior, making transactions more integrated and less reliant on traditional web-based checkout processes.

Collison reflects on the broader industry landscape, mentioning the competitive race among tech giants like Google, Microsoft, and emerging startups in AI and coding tools. He notes that the pace of innovation is accelerating, likening it to the tech race of the 1980s, with smaller startups challenging established players. While acknowledging the productivity gains from AI, he cautions that increased expectations and competition may lead to faster product releases and heightened industry rivalry. He also discusses the importance of security, reliability, and global expansion, especially amid geopolitical uncertainties like tariffs and trade policies.

Finally, Collison shares insights from his admiration for Warren Buffett and Charlie Munger, emphasizing the importance of long-term thinking, preparedness, and patience in business. He explains that Stripe’s growth strategy is focused on building durable, scalable products that can adapt to future technological shifts, much like Berkshire Hathaway’s approach to value investing. Collison concludes by reaffirming Stripe’s commitment to innovation and stability, noting that their current focus is on leveraging AI and other tailwinds to enhance their offerings and sustain long-term growth, rather than rushing toward an IPO.