Klarna's AI does work of 700 agents, projects $40M boost

Klarna’s numbers aren’t subtle. Two months after going live, the fintech’s OpenAI-powered assistant has handled 2.3 million customer service chats — roughly two-thirds of the company’s total volume. That’s the equivalent of 700 full-time employees. The kicker? It’s not just matching human performance; it’s beating it on accuracy, with repeat inquiries dropping 25%. Customers who used to wait 11 minutes for resolution now wrap things up in under two minutes.

CEO Sebastian Siemiatkowski didn’t wait around when ChatGPT launched in November 2022. Klarna became the first European company and first fintech globally to ship a ChatGPT plugin, letting shoppers describe what they want in plain language and get back product links with price comparisons. Martin Elwin, Senior Director of Engineering, framed the effort as addressing existing headaches while pushing toward a fundamentally different shopping experience. The vision is straightforward: ask for sneaker recommendations under $150 and refine the search conversationally, rather than doom-scrolling through thousands of near-identical product pages.

The internal adoption numbers tell an equally aggressive story. Klarna rolled out ChatGPT Enterprise across the entire organization, and 90% of employees now use generative AI tools daily. Communications, Marketing, and Legal teams clock adoption rates of 93%, 88%, and 86% respectively. Use cases run the gamut from drafting external messaging to building software and streamlining legal workflows.

Siemiatkowski estimates the AI assistant alone will add $40 million in profit improvement during 2024. Yet he’s not doing victory laps. In his statement, he pivoted to a broader warning about the societal disruption AI signals, urging politicians and the public to treat this moment with “considerate, informed and steady stewardship.” That’s the tension at the heart of this rollout: Klarna is sprinting toward AI-driven efficiency while its CEO is simultaneously telling everyone to slow down and think hard about what comes next.