Layoffs and AI

Unlock Editor’s Digest for free

FT editor Roula Khalaf selects her favorite stories in this weekly newsletter.

Good day. Bitcoin and gold are down 11 percent and 8 percent, respectively, from their October highs. The dollar is based, not debased, right? Have fun staying poor, skeptical fiduciaries! Also, send us an email: unhedged@ft.com.

Layoffs and AI

Well, this doesn’t sound very good. From my colleagues at the Financial Times:

Corporate workers are fighting for a shrinking pool of jobs as mass layoffs and a slowdown in hiring stoke fears of a “white-collar recession,” amid growing pressure from tariffs, slowing economic growth and artificial intelligence.

Last week, companies including Amazon, Paramount, United Parcel Service and Target announced they will collectively cut 31,800 office jobs, with some directly citing plans to use AI to reduce their workforces. Others cited rising costs from President Donald Trump’s tariff regime and falling sales as consumers cut back on spending.

There have been a noticeable number of high-profile layoff announcements lately, and journalists have been busy looking at them from one angle or another. Two related points dominate the coverage: AI and young people. It is argued that the fact that big tech companies (namely Amazon and Meta) are among the companies announcing layoffs makes the job losses more sinister: a glimpse into the econotech future. “Those big technology companies are the first to adopt [of AI] so we’re likely to see the impact there first,” a source said in the Financial Times story. Secondly, young people are said to be bearing the brunt, possibly because entry-level jobs are the easiest for AI to replace (a very good idea). history in the Wall Street Journal went into detail on this aspect).

Sometimes a supposed economic “trend” turns out to be simply made up: an issue becomes so socially or culturally relevant that observers only see what they have been primed to see. That doesn’t seem to be the case here: at least there seems to be an unusually high number of layoff announcements, according to the usual source on this topic, the monthly newsletter report from “executive relocation” firm Challenger, Gray and Christmas. From the latest edition:

So far this year, companies have announced 946,426 job cuts, the largest to date since 2020, when 2,082,262 were announced. It’s 55 percent more than the 609,242 job cuts announced during the first three quarters of last year and 24 percent more than the total of 761,358 for all of 2024. The 2025 year-to-date total is the fifth highest in the 36 years that Challenger has reported.

But more layoff announcements don’t mean more layoffs. And there is very good data to suggest that the number of people losing their jobs is not increasing significantly: new unemployment claims, which are collected at the state level and therefore (mostly) unaffected by the shutdown. Below is the trend for them. The compilation and extrapolation of state figures available since September comes from JPMorgan:

Weekly new claims have remained fairly stable for three years, mostly fluctuating in a band between 200,000 and 260,000. And the picture may be improving: Employ America’s Skanda Amarnath notes that more states are reporting declining claims levels than increases compared to a year ago, and this ratio is on an improving trend.

This panorama is more or less reinforced by the layoffs part of the Job Offers and Labor Rotation Survey, which we have until August. As many people have pointed out recently, the biggest problem doesn’t seem to be more layoffs, but less hiring:

Having said all this, it is clear that the weakening demand for labor is manifesting itself in a very weak labor market for young people. A long-term graph of unemployment among 20- to 24-year-olds shows that unemployment in this group is at levels previously seen at the start of recessions:

This brings us to murkier questions about whether AI is having an impact on the job market, particularly at the entry level. I’m not sure if it’s possible to answer this question right now, and I’m skeptical of corporate announcements that cite AI as a reason to cut jobs. At this point, how would companies know, outside of very specific tasks, how much sustained productivity improvement they can expect from AI? CNBC’s Michael Santoli commented in Bluesky that

We’re simply in a mode where corporate managers are moving beyond over-hiring, increasing margins, navigating large technology spending commitments, and conveying to investors that they are forward-thinking and lean-minded.

This makes sense to me. In the case of Big Tech, those companies have been in on-again, off-again job cuts mode since the post-pandemic hiring frenzy. And the markets, as we have argued recently, have been quite demanding about the profits and cash flow of large companies; Corporate leaders will be aware of this. Attributing layoffs to AI sounds better than saying “we need to keep margins high, so we’re laying off some low-performing workers” and politically safer than saying “Trump’s unpredictable tariff policy means we’re hiring fewer young people.”

a good read

Trump’s positive impact.

Podcast without FT coverage

Can’t get enough of Unhedged? Hear our new podcastfor a 15-minute dive into the latest market news and financial headlines, twice a week. Catch up on previous editions of the newsletter here.