Revolut Builds AI Models From 40 Million Daily Transactions
Revolut is developing proprietary AI models using data from around 30–40 million daily transactions, CEO Nik Storonsky said at a technology conference in Turin. The fintech has not disclosed specific commercial targets or expected AI revenue.
Storonsky also said Revolut would favor a primary U.S. stock-market listing if it goes public. An IPO is considered unlikely before 2028, and no date or exchange has been confirmed. A secondary share sale valued the company at $115 billion this year, up from about $45 billion in 2024.
Revolut is also pursuing U.S. banking expansion. It received preliminary approval for a national banking license in September, but further regulatory steps are required before it can offer all planned services. Its established businesses generated about $6 billion in revenue during 2025. The Revolut AI strategy and potential U.S. IPO underline the company’s ambition to compete beyond European digital banking.
Neutral
The announcement has no clear, immediate effect on cryptocurrency prices or market stability. Revolut’s AI development and possible U.S. IPO are company-specific developments, and the article reports no new crypto product, token investment, or change in digital-asset policy. The preliminary U.S. banking approval is also not a full authorization, limiting the near-term impact.
In the short term, traders may view the news as a modest signal of continued investment in financial technology, but it is unlikely to move major crypto assets without a broader risk-on or risk-off catalyst. Past announcements about banks and fintech firms adopting AI have generally affected sentiment more than prices directly; market reaction tends to depend on whether they lead to concrete product launches, revenue, or regulatory changes.
Over the longer term, proprietary AI and a wider banking footprint could help Revolut improve services and compete with established financial firms. If that eventually includes expanded crypto access or infrastructure, it could matter for adoption and trading flows. No such plans are stated here, however, so a neutral classification is most appropriate.