American Express to Cut About 5% of Its Global Workforce in AI-Driven Overhaul

American Express said on Friday that it would eliminate about five per cent of its global workforce — roughly three and a half thousand jobs — as the payments giant reorganises itself to prepare for a deeper embrace of artificial intelligence, even as it reported one of the strongest financial quarters in its history.

Chief executive Stephen Squeri announced the cuts on the company’s third-quarter earnings call with analysts on Friday morning, describing the move as a deliberate reshaping of the organisation rather than a response to weakness in the business. The company will book a US$336 million pre-tax restructuring charge tied to the job cuts, according to reporting by CNBC on the earnings call.

“We’ve got good momentum, but to accelerate innovation and win over the next few years, we needed to become a more agile and efficient organisation,” Squeri told analysts, according to CNBC’s reporting. “When completed, we believe this exercise will help us operate and execute more effectively and give us greater capacity to invest in our business priorities.”

The decisive phrase, however, came from the company’s chief financial officer. Christophe Le Caillec told CNBC in an interview that the restructuring was intended to make American Express “leaner, more agile, and also to prepare for AI”.

“You don’t want to automate processes that are not optimally structured,” Le Caillec told the network. “So what we’re doing here is to reduce layers and increase spans of control, so that we can automate and take advantage of AI in the coming quarters and the coming years.”

The company had nearly 77,000 employees at the end of 2025, according to its own filings, which puts the reduction at roughly 3,800 positions, according to Fast Company, which confirmed the layoffs with the company. Bloomberg, citing a figure given by Squeri in a separate interview, put the total at about 3,500 of roughly 78,000 roles. American Express has not said which teams, regions or business units will bear the heaviest losses, nor how quickly the cuts will be completed.

A record quarter — and a restructuring charge on top of it

The cuts arrive wrapped inside a financial performance that most companies would envy. American Express reported record third-quarter revenue of US$20.1 billion, up nine per cent on a year earlier, and net income of US$3.7 billion, up 28 per cent, according to the company’s earnings release filed with the US Securities and Exchange Commission on October 23.

Diluted earnings per share came in at US$5.42. On an adjusted basis, earnings of US$4.68 a share topped the US$4.48 average estimate of analysts compiled by LSEG. Billed business — the total value of transactions charged on the company’s cards — reached US$454 billion, up eight per cent. The company also raised its full-year 2026 earnings guidance to a range of US$18.30 to US$18.60 per share, up from its previous target of US$17.30 to US$17.90.

The US$336 million restructuring charge works out at US$253 million after tax, according to the SEC filing, and is described in the filing as “restructuring related to efficiency initiatives”.

That juxtaposition — record profits on the one hand, thousands of jobs cut on the other — is becoming a defining pattern of the AI era’s corporate playbook. The cuts are not being framed as a rescue measure. They are being framed as a strategy for staying ahead.

What the company says it wants AI to do

Squeri reportedly devoted roughly ten minutes of his prepared remarks to artificial intelligence, according to coverage of the call. The focus, he said, was on using AI to personalise experiences and offers for cardholders — the kind of customisation that a premium card brand sells as its core product.

But he also drew a hard line around how far AI agents will be allowed to go inside the company’s systems. American Express is “not going to let assistants run rogue through our system,” Squeri said, according to reporting on the call. He added that the company would only permit agents to access its systems through approved channels — specifically API or MCP channels, the two standard ways of securely connecting outside software platforms to a company’s infrastructure.

“We need to protect our customers and merchants,” he said, according to the reporting, adding that meeting customers’ desire to use AI agents has to happen within guardrails.

The language is carefully calibrated: enthusiastic about the technology, anxious about its risks, and explicit that the company intends to be the one deciding how it is deployed.

Part of a wider wave in finance

American Express is not acting in isolation. A day earlier, the British banking giant Barclays confirmed plans to cut about 5,000 jobs — also roughly five per cent of its staff — as part of a technology-driven cost initiative that it linked to artificial intelligence, according to CNBC’s reporting. The two announcements, 24 hours apart, amount to nearly nine thousand financial-sector jobs cut with AI cited as a driving factor.

They join a growing list. Block, Coinbase and Standard Chartered have all cited artificial intelligence as a key driver of workforce reductions this year, according to reporting by Wealth Professional, part of a broader group of more than 40 companies that have announced layoffs in 2026 linked to AI, according to Business Insider’s tracking.

The financial sector is emerging as the clearest early test of a particular theory of AI adoption: that the first and largest gains from automation will come not in factories or warehouses, but in back offices — compliance, customer service, risk analysis, document processing — where armies of staff perform repetitive, rules-based work on screens all day.

Analysis: Why It Matters

The “restructure first, automate later” doctrine

The most important sentence in this story may be the one the chief financial officer gave to CNBC: “You don’t want to automate processes that are not optimally structured.” That is not a company saying AI is replacing these workers today. It is a company saying it wants its organisation flat enough that AI can absorb the work tomorrow.

This is a meaningful distinction, and it is worth taking seriously rather than dismissing as corporate euphemism. If American Express were simply swapping humans for chatbots, it would be rolling out the automation first. Instead, it is removing management layers now — the people whose job is to coordinate, approve and oversee — on the assumption that automated systems will handle more of that coordination soon.

For workers, the practical difference is thin: the job is gone either way. But for anyone trying to read where this is heading, the sequencing matters. The company is placing a bet, at the cost of thousands of careers, that the automation will arrive on schedule. If it does not arrive as quickly as expected, the company will have cut its coordination capacity with nothing to replace it.

Cutting from strength, not from weakness

There was a time when mass layoffs were an admission of failure — a sign that a company had misjudged its market, overextended itself, or been outmanoeuvred. That logic no longer holds. American Express has just posted record revenue, rising profits, an earnings beat and raised guidance, and it is cutting five per cent of its workforce in the same breath.

The signal this sends to every other profitable company is unmistakable: you do not need a crisis to cut. You need a strategy document that says AI will let you do more with fewer people, and a board willing to act on it while times are good rather than after they turn bad.

That logic will be tested in the coming quarters. Investors rewarded the move today. Whether the remaining 95 per cent of the workforce — and the customers who depend on their judgement — experience an organisation that is genuinely more agile, or merely thinner, is a question no earnings call can answer.

Finance is AI’s proving ground — and its warning shot

Banks and payments companies are in the vanguard of AI-driven job cuts for a structural reason. Their workforces are unusually exposed: large, well-paid, concentrated in offices, and performing tasks — document review, transaction monitoring, customer authentication, basic underwriting — that machine-learning systems handle increasingly well.

What happens in finance this year is likely a preview of what happens in insurance, professional services, logistics administration and large parts of government bureaucracy next year and the year after. The Amex and Barclays announcements, a day apart, are not two isolated decisions. They are the first two data points in a trend line that every white-collar worker should be watching.

The counter-argument deserves its due: the same firms are also hiring AI specialists, and some of the “capacity to invest” Squeri mentioned will go into roles that did not exist five years ago. But the net arithmetic so far is unambiguous, and it points in one direction.

The disclosure gap

One notable feature of the announcement is what it does not say. American Express has given no breakdown of which regions or business units will be hit, and no timeline for when the cuts will be completed. For thousands of employees, Friday’s announcement is the beginning of a period of uncertainty rather than its end.

There is also a transparency question that will follow this story into earnings season. Companies are happy to cite AI as the reason for cuts — it signals modernity to investors — but there is little independent verification of how much of any given reduction is truly driven by automation versus ordinary cost discipline wearing an AI badge. Until firms report AI-specific productivity metrics alongside their headcount numbers, the claims will remain impossible to audit.

The agent guardrails tell their own story

Squeri’s warning that American Express will “not let assistants run rogue” through its systems is revealing in a different way. It acknowledges, from inside one of the world’s largest financial institutions, that AI agents — software that can take actions rather than just answer questions — are now capable enough that a bank must actively defend its infrastructure against them.

That is a striking admission of where the technology stands. The company wants customers to use AI agents, wants to personalise offers with AI, and wants to automate its own operations — but through approved API and MCP channels only, with the bank holding the keys. Control, not capability, is now the battleground. Whoever sets the terms on which agents are allowed to act inside financial systems will shape the next decade of the industry.

What to Watch Next

1. Which units take the hit: when American Express files its annual report, the regional and segment breakdowns may reveal whether the cuts concentrate in customer service, technology operations or corporate functions — each telling a different story about what AI is actually replacing.
2. The Barclays comparison: with two major financial firms cutting five per cent within 24 hours, analysts will watch whether rivals — JPMorgan, Citigroup, HSBC — follow with their own AI-linked restructuring announcements in the coming weeks.
3. Headcount versus hiring: track whether Amex’s total headcount actually falls, or whether AI-specialist hiring offsets the cuts — the net figure will show whether this is displacement or transformation.
4. The $336 million test: management has promised greater agility and execution capacity. The restructuring charge gives investors a concrete number against which to measure the promised efficiency gains in 2027 results.
5. Regulatory scrutiny: as AI-driven cuts spread through finance, watch for policymakers and unions to demand clearer disclosure of how automation figures into workforce decisions — particularly in jurisdictions with strong worker-protection rules.

Sources

• Fast Company — “American Express layoffs today: Amex slashes jobs in AI push”
• Canadian HR Reporter — “Amex to cut 5 per cent of workforce ‘to prepare for AI’” (reporting CNBC and Bloomberg)
• Marketscale — “American Express will cut about 3,500 jobs while raising its 2026 forecast” (citing CNBC, Bloomberg, SEC filing and LSEG)
• Wealth Professional — “Payments giant makes layoffs citing AI”