AI has been the top reason US employers cite for layoffs, five months running now. Challenger, Gray & Christmas put July 2026's AI layoffs at 10,970, a third of everything announced that month, and the year's running total has already cleared 112,000. A few years back, blaming a layoff on a chatbot would have sounded evasive. Now it's just what goes in the press release.
Not all of it holds up once you look closer, though. Oxford Economics ran the numbers and found AI explained only 4.5% of the roughly 55,000 US job losses blamed on it through 2025, with weak demand and ordinary cost-cutting doing the rest. Some CEOs describe AI replacing jobs in plain language. Others deny AI had any role in a round of cuts, then predict it'll shrink their workforce within a few years.
The line between an AI layoff and an AI excuse
A layoff earns the AI label when a company can point to a specific piece of AI automation that took someone's job, and the headcount follows. It's AI-adjacent when AI gets mentioned next to a restructuring that was probably coming anyway. The IMF has put a number on the exposure side of this: roughly 60% of jobs in advanced economies carry meaningful AI exposure. Research on AI and employment agrees on one caveat: exposure isn't elimination.
The companies that named AI directly
These are the ones where a CEO tied AI directly to a round of cuts, not just a vague nod to efficiency.
Salesforce
In September 2025, Marc Benioff stated that with the help of Agentforce, an AI agent technology offered by Salesforce, the company was able to reduce its sizeable customer service force of approximately 9,000 to just around 5,000, effectively reducing the workforce by 4,000 employees. He put it bluntly: he needs "less heads" now that agents handle most of the support load.
Block
Block cut roughly 40% of its staff in a single move, taking headcount from over 10,000 to under 6,000, and Jack Dorsey explained it in a shareholder letter, not a memo: a smaller team using the AI tools Block is building "can do more and do it better," he wrote, pointing to recent coding models as the trigger. The stock rose about 22% on the announcement.
Coinbase
The exchange fired nearly 700 workers in May 2026, which was around 14% of its workforce. This layoff move had been linked directly by its chief executive Brian Armstrong with the speed at which artificial intelligence is disrupting the company. It now runs on what it calls "AI-native" pods, with pure managers mostly gone, replaced by "player-coaches" who still work hands-on. Coinbase expects $50 to $60 million in restructuring charges, mostly severance, while still growing its AI engineering headcount.
Kraken's parent company
Payward, the maker of Kraken, is not as vocal on the subject. Payward let go of around 150 employees in May 2026 following AI adoption, representing about 5% of its workforce of 3,000 people, as part of a move to optimise ahead of an IPO which has now been postponed to 2027 due to falling prices.
Dune Analytics
But Dune chose a new strategy. It reduced its workforce by 25% in May 2026, not to survive, but to strengthen its bets on AI agents and institutional on-chain data. CEO Fredrik Haga pointed to Dune MCP, a server letting AI agents query its data warehouse directly, as the reason. Unlike most names on this list, Dune says the balance sheet was never the problem. Unlike many of the others listed here, Dune insists the problem wasn’t the balance sheet.
IBM
IBM's version got exaggerated along the way. CEO Arvind Krishna says AI automated a couple hundred of the company's HR roles, with an internal tool, AskHR, handling most routine requests. That's a long way from the bigger numbers that circulated in some 2025 headlines. Krishna explains how the savings were put back into the company to hire more engineers and salesmen despite fewer HR workers.
Where AI is one reason among several
Not every executive wants credit for this: some downplay AI's role even when the timing looks suspicious, others are dealing with pressures big enough on their own to explain the cuts.
Amazon
In October 2025, Amazon announced cuts of around 14,000 jobs from its corporate workforce, and Andy Jassy, CEO of Amazon, stated that AI was not responsible for the cuts because the move "was not really financially driven, it's not even really AI-driven, not right now, at least," he said, blaming years of added layers instead. adding that there were many years of layers built up. This contradictory statement is hard to reconcile with Amazon’s own statements regarding how gains made by AI technology will reduce its corporate workforce over the coming few years. In January 2026, a further 16,000 jobs were cut.
Meta's Reality Labs
Meta has laid off several people from Reality Labs and elsewhere starting in early 2026, shifting funds to AI technology development. The department has incurred losses ever since it was formed by Meta, trying to create the metaverse that failed to materialise. It sounds less like AI taking their jobs away and more like AI feeding on the funds allocated to them. Meta increased its 2026 AI budget forecasts by as much as $10 billion in the same period.
Ethereum Foundation
The version that the Ethereum Foundation has adopted is a messier approach. This organisation had to lay off 20% of its workforce, about 54 employees, in June 2026, at the same time reducing their budget by 40% and losing nine top managers since January. The remaining employees have been reallocated into five clusters, including one cluster which revolves around AI agents in the protocol.
OP Labs
OP Labs trimmed about 20 positions in March 2026, and according to their CEO, Jing Wang, it had absolutely nothing to do with money or AI. He said it was just about doing “fewer things well” and minimising the overhead of coordination. It comes at a time when a competing chain is switching over to its own tech stack, a real blow to Optimism that has nothing to do with automation.
Major 2026 announcements at a glance
Put next to each other, these ten companies look less like one trend than ten separate bets, each with its own justification.
Which jobs are disappearing, and where AI is hiring
AI job cuts hit customer support and back-office work first, the repetitive kind large language models already handle well. Even the media industry was not spared as the news portal Business Insider fired 21% of its workforce in 2025 as it goes "all-in on AI," a decision described by the union as a "shift away from journalism." On the flip side, AI engineer is currently the fastest-growing job title in the US, with job postings increasing by 143% from last year. This does not average out, though, because the World Economic Forum predicts an estimated creation of 170 million jobs while displacing 92 million jobs by 2030.
The savings show up unevenly
Some of it is reflected in the numbers that investors see. Salesforce notes a 17% reduction in support expenses as a result of the introduction of Agentforce. Block’s stock price surged on the announcement of its layoffs, while Coinbase made sure to include its charge in its guidance to avoid the element of surprise. It's not all clean wins, though. Klarna is the story everyone brings up now: after cutting roughly 700 customer service jobs for an OpenAI-built assistant, its CEO admitted cost had mattered more than quality and started rehiring once complaints piled up.
The rehiring problem
This is not an isolated case of an embarrassing situation. According to Gartner, half of the firms that will reduce their customer service employees due to the implementation of AI will hire new workers for the exact position by 2027. However, they may change the position name. This does not imply that the technology is doomed to fail. It simply means that the first wave of reductions occurred faster than the technology could catch up, and now firms use a hybrid approach. It remains to be seen whether this trend repeats itself elsewhere.
The lean-startup argument behind these cuts
You can see the same logic in how new companies get built. The ten largest AI-native startups run with about 24 employees on average, pulling in roughly $3.48 million in revenue per employee, against around $610,000 at a typical established SaaS company. That gap is the real argument for AI-linked layoffs at bigger, older companies: a small team with the right tools can compete with a division ten times its size, and boards have noticed. It's also why the next round of cuts probably won't wait for a downturn.
Tanya Petrusenko
Disclaimer: All materials on this site are for informational purposes only. None of the material should be interpreted as investment advice. Please note that, despite the nature of much of the material created and hosted on this website, HODL FM operates as a media and informational platform, not a provider of financial advisory services. The opinions of authors and other contributors are their own and should not be taken as financial advice. If you require advice, HODL FM strongly recommends contacting a qualified industry professional.
