Klarna’s CEO has predicted that a recession could be around the corner as companies across the globe—including his own—reduce the headcount of well-paid, white collar jobs and replace them with AI.

Sebastian Siemiatkowski, the boss of the Swedish Buy Now, Pay Later group is once again sounding a pessimistic tone on AI’s impact on the workforce. But as he embraces the potential positive effects of AI on his own bottom line, he may have to contest with the negative fallout of a company that has flirted with growing credit losses in the last year.

While he admitted that “making future statements about macroeconomics is like horoscopes,” Siemiatkowski’s well-documented feelings about AI’s impact on the labor market leave him making a pessimistic prediction about the economy.

“My suspicion… is that there will be an implication for white collar jobs. And when that happens, that usually leads to at least a recession in the short term. And I think, unfortunately, I don’t see how we could avoid that with what’s happening from a technology perspective,” Siemiatkowski said on the Times Tech Podcast.

Siematkowski has long warned of the disruptive nature of AI on the labor market, using his experience of shiting recruiting practices at Klarna to support his argument that it will replace roles.

He told the podcast that the company’s headcount had fallen from 5,500 people to 3,000 in the space of two years. Speaking in August last year, Siematkowski said his ambition was to eventually reduce that figure to 2,000 through workplace norms like attrition rather than by engaging in layoffs.

In February last year, Klarna announced that its AI chatbot was doing the work of 700 customer service staff, previously a role filled by customer service agents working for the French agency, Teleperformance.

While Siemiatkowski has faced criticism for his willingness to talk about AI’s disruptive potential, he indicated he felt it was more of a duty to be frank about the technology.

“Many people in the tech industry, especially CEOs, tend to downplay the consequences of AI on jobs, white collar jobs in particular, I don’t want to be one of them.”

Indeed, Siemiatkowski implied that if he added up the number of employees of CEOs who had called him to ask about making “efficiencies,” that figure in itself would make for a seismic economic event.

Recession indicator?

An AI-induced recession would combine a number of brewing themes for the Swedish tech group. Siemiatkowski’s comments come as the group reported widening credit losses, which rose by 17% to $136 million last year.

Siemiatkowski explained the losses as a result of the group taking on more customers, naturally leading to a rise in defaults. On a relative basis, the percentage increase in defaults was small, Siemiatkowski said.

The Swede added that because Klarna customers’ average indebtedness was £100, they were more likely to pay back their loans compared with typical credit card debt of what he said was £5,000. The typical U.K. credit card holder has an outstanding credit balance closer to around £1,800, while in the U.S., the average is about $6,300.

Regardless of the variance, Siemiatkowski says the difference means customers are more likely to pay off their Klarna debts.

“We are very unsensitive to macroeconomic shifts. We can still see them, but they’re much less profound than if you’re a big bank, you have tons of mortgages. And for people to really increase losses, credit losses, what has to happen is people have to lose jobs.”

Despite that, predictions of mass layoffs amongst white collar workers could inform higher risk for the company’s credit business.

While there wasn’t any sign of a recession currently, Siemiatkowski did observe falling consumer sentiment, which would impact spending.

Siemiatkowski’s views on AI in the labor force have evolved over time. Speaking to Bloomberg in May, Siemiatowski was reported to have said the company was embarking on a recruitment drive, contrary to his previous statements about a workforce reduction.

Speaking with the Times, Siemiatkowski clarified that the company needed different types of workers to handle more complex customer service requests.

“When we started applying AI in our customer service, we realized that there will be a higher value to human connection,” he said.

This story was originally featured on Fortune.com