Klarna CEO warns AI could cut jobs, firms unprepared

Klarna CEO Sebastian Siemiatkowski warned that AI could eliminate many knowledge-based jobs, including in banking and finance.

Klarna, a Swedish fintech known for its buy-now-pay-later service, has heavily invested in AI to cut costs and boost efficiency.

The company reduced its workforce from 7,400 to about 3,000 after the fintech boom ended and paused hiring for over a year.

In 2025, Klarna hired more customer service staff to ensure users can reach human agents, while using AI selectively in underwriting.

Siemiatkowski said AI may reduce profits for banks and software firms as it handles tasks efficiently, and Klarna plans to expand its services in the UK.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

Klarna’s AI efficiency gains lack operational transparency on quality metrics

Reports say Klarna’s AI handles two thirds of chats and cut resolution from 11 to 2 minutes 1. The company has not shared data on first‑contact resolution (issues solved in the first interaction), customer satisfaction, or complaint trends.
Sources list a 4.1/5 Trustpilot (a consumer review site) rating plus over 900 Better Business Bureau (BBB, a US nonprofit that tracks consumer complaints) complaints over three years on refunds or billing 1.
Klarna CEO Sebastian Siemiatkowski said cost was “a too predominant evaluation factor” 2. The company rehired human agents in 2025 2, which underlines that complex disputes still need judgment.

EU AI Act compliance gaps create opportunity for credit decisioning infrastructure vendors

The European Union’s AI Act (a comprehensive regulation for AI systems) marks credit scoring of natural persons (individual consumers) as high risk, with risk management, explainability (the ability to clearly describe how a model reaches its outputs), plus human oversight before rollout 3.
Financial institutions face AI Act applicability from August 2026 3. Buyers want credit decisioning infrastructure (the software plus controls used to evaluate credit approval) with built-in guardrails (controls that prevent misuse or errors), audit trails, plus documentation that meets transparency rules 4 5.
Business-to-Business (B2B) software providers can map AI Act obligations into turnkey solutions (ready to deploy products) 6. Packages that cover risk assessments, bias mitigation, plus explainability can win banks and fintechs (financial technology companies) that need compliant credit models yet lack governance 6.