Stripe has announced the launch of what it describes as the world´s first Artificial Intelligence foundation model specifically designed for the payments industry. The new model, trained on tens of billions of transaction records, is engineered to capture hundreds of data signals about each payment—signals that Stripe says surpass those detected by previous models. By deploying this technology across its payments suite, Stripe aims to significantly enhance payment performance, security, and fraud prevention.
According to Stripe, prior machine learning efforts saw an 80 per cent reduction in card testing over two years. With the new foundation model, the company reports a 64 per cent increase in detection rates for cyberattacks on large businesses almost immediately after deployment. These advances highlight the model´s capability to proactively counter evolving payment fraud threats and improve overall transaction reliability for Stripe´s business customers.
In tandem with the Artificial Intelligence launch, Stripe introduced Stablecoin Financial Accounts, a global money management solution allowing businesses in 101 countries to hold, receive, and send stablecoin balances. This feature is designed to support entrepreneurs in regions with volatile currencies, enabling them to hedge against inflation and participate more effectively in the global economy. Stripe´s expansion into multicurrency management also includes the rollout of features enabling businesses to store, convert, and spend funds in multiple currencies, helping multinational clients avoid frequent foreign exchange fees. Initially available to firms in the US and UK, these tools will reach the Eurozone later in the year. Stripe´s president of product and business, Will Gaybrick, stated: ´We’re building programmable financial services to make money as easy to manipulate and manage with code as data is,´ underscoring the company’s commitment to accelerating user growth through the broad integration of Artificial Intelligence and stablecoins.
