Klarna went further with AI automation than almost any business its size. It was right to do so. Then it overcorrected. That distinction matters for every business leader currently making decisions about AI and their workforce.

In early 2024, Klarna's AI customer service assistant handled 2.3 million conversations in its first month. It resolved two-thirds of customer inquiries. It cut resolution times from an average of 11 minutes to under two minutes. The company's CEO, Sebastian Siemiatkowski, went public with the results. The automation was working.

But by May 2025, Siemiatkowski had a different message. "We went too far," he told reporters. "We focused too much on cost. The result was lower quality, and that's not sustainable." Klarna was hiring human customer service agents again.

The AI had not failed. The decision-making around the AI had. The company had reduced its human workforce faster than it could verify quality was being maintained. Customer satisfaction declined. The company course-corrected — but the headlines stuck.

There is a precise lesson buried in this sequence, and it is not the one being repeated in most coverage of the story.

The lesson is not that AI cannot handle customer service. It clearly can, at scale, and the Klarna deployment continued running. The lesson is that automation decisions made purely on cost produce cost-shaped outcomes. When you optimize for what AI can do without maintaining human oversight at the right points, you lose the signal that tells you something has gone wrong.

The businesses getting this right are not removing humans from customer-facing workflows. They are redeploying them. AI handles volume, pattern recognition, and first-line resolution. Humans handle exceptions, escalations, and the cases where judgment and empathy are the actual product. The combination produces better outcomes than either approach alone.

This is exactly where tools like Viktor operate. Not as a replacement for people, but as the layer that handles structured, repeatable work so that people can focus on the work that requires human judgment.

Viktor lives inside Slack and Microsoft Teams. You @mention it in a thread the same way you would ask a colleague. The output — a PDF, a report, a task created in your CRM, an email drafted in Gmail — lands where it should land.

The practical applications are straightforward. An account manager who previously spent three hours preparing client briefings before calls can now prepare in 20 minutes, with Viktor pulling together the relevant data, drafting the briefing, and flagging anything that needs attention. The account manager still makes the call. Still reads the room. Still makes the judgment that converts a prospect. The preparatory work is automated; the human work is elevated.

This is not a theoretical workflow. It is the model that separates businesses that benefit from AI from those that end up rehiring after a painful public correction.

Klarna's mistake was not trusting AI. Its mistake was assuming that cost reduction was a sufficient measure of success in a customer-facing function where quality is visible and immediate. The businesses learning from this are the ones building AI into their operations as a force multiplier for their people — not a substitute for them.

If you are evaluating AI tools for business use, security and compliance are not optional considerations. Viktor holds SOC 2 Type II certification and complies with GDPR, CCPA, and CASA Tier 3. Credentials are stored in an encrypted vault and are never used to train models. Sensitive actions can require Slack-based human approval before execution. Full details are at viktor.com/security.

You get $100 of free credits to begin. No time limit, no commitment. That's enough to do real work and see what Viktor can actually do before you spend a penny. There's also $50 off your first bill.

You must use this exact link to receive both benefits: AIThatDelivers.com

Disclosure: Some links in this article are affiliate links. If you choose to get started with Viktor using the links provided, I may receive a commission — at no additional cost to you. I only recommend tools I use and believe in.