The shift started with Klarna’s early work with OpenAI, first around shopping discovery and later through its AI customer service assistant. That assistant handled 2.3 million conversations in its first month, covering two-thirds of customer service chats while maintaining human-level satisfaction scores. For a business with more than 100 million users and high transaction frequency, this is not just a service improvement. It changes the cost base.

The more important layer is credit. Klarna’s model depends on making fast, accurate decisions at checkout, often on short-duration loans and thin transaction windows. AI allows Klarna to analyse behavioural and transactional signals in real time, improving approval quality while managing risk. The reduction in credit loss rates from 3.6% to 0.63% over four years shows how underwriting has become a major part of the AI advantage, not just an operational use case.

AI is also strengthening Klarna’s commerce layer. By understanding intent at the SKU level, Klarna can surface relevant products, offers, and locations before the customer reaches checkout. That turns Klarna from a payment option into a discovery and conversion tool for merchants, supported by click-through rates that are 8x higher than leading competitors.

The internal productivity angle completes the picture. With 96% of employees using AI daily, Klarna is applying automation across code, data, legal, marketing, and operations. The result is a business where AI supports both front-end growth and back-end efficiency.

Klarna’s AI advantage is therefore not one product or one assistant. It is the way AI is being embedded into the core mechanics of the business: deciding who gets credit, serving customers at scale, improving merchant conversion, and reducing operating intensity as the company grows.