Games Contact us
Select Page
⚡ TL;DR
On October 23, 2026, American Express said it will cut about 5% of its global workforce, roughly 3,500 of about 78,000 employees, with a $336 million pretax restructuring charge, while raising its 2026 earnings guidance the same day. Management framed the move as flattening management layers and preparing for AI, not as replacing staff with software. For HR leaders, the episode is a case study in how to read, plan for and communicate “AI-readiness” restructurings.

American Express posted a strong quarter, lifted its outlook and announced thousands of job cuts in the same release, a combination that is becoming a recognisable pattern in financial services. CEO Stephen Squeri disclosed the plan on the post-earnings call. Third-quarter revenue rose 9% to $20.1 billion, and the company raised its 2026 GAAP earnings-per-share guidance to a range of $18.30 to $18.60, up from $17.30 to $17.90. Investors, however, focused on the cuts and on slower spending growth: shares fell about 3.7% to roughly $313 on the day.

Key Takeaways

  • About 3,500 roles, around 5% of the workforce, are being eliminated, according to CNBC, Reuters and Bloomberg reporting.
  • The company recorded a $336 million pretax restructuring charge, reportedly offset in part by a $982 million gain on selling a stake in Global Business Travel Group.
  • Management said the aim is a leaner, more agile company that also prepares for AI. It has not said AI is replacing the eliminated roles.
  • Amex did not disclose which units or countries are affected, the timeline, or the expected savings.
  • Barclays announced about 5,000 job cuts a day earlier, also citing AI, which points to a sector-wide trend.

The numbers behind the announcement

The headline figure is 3,500 jobs. Most outlets calculate that as 5% of a workforce of about 78,000. One tracking site lists a smaller cut of 4.6% against a base of 76,800, a reminder that workforce baselines differ depending on whether contractors, part-time employees and recent hires are counted. For planning purposes the difference is minor; for anyone benchmarking the cut against peers, it is worth using the company’s own figures once they appear in regulatory filings.

The financial structure is more telling than the headcount. The $336 million restructuring charge works out to roughly $96,000 per eliminated role, a figure that plausibly covers severance, benefits continuation, outplacement and associated costs. That is an average implied by two reported numbers, not a disclosed per-employee figure, and it will vary widely by country and seniority. The reported $982 million gain from the sale of the Global Business Travel Group stake is larger than the charge, which helps explain why the company could absorb the cost without disturbing its adjusted earnings.

“Prepare for AI”: what the language does and does not say

The phrase that dominated coverage was the CFO’s description of the goal: a company that is “leaner, more agile, and also to prepare for AI.” Squeri described the practical mechanics in terms of removing management layers and widening spans of control, so the company can automate more later. Headlines, understandably, turned this into “Amex cuts jobs to make way for AI.” The company’s own statements are more careful. It has not attributed the cuts to a deployed system, and it has not said that software is taking over the work of the people leaving.

That distinction matters for HR for three reasons.

  1. It shapes how remaining employees interpret the message. If the explanation is “AI replaces roles,” every team worries it is next. If the explanation is “we are simplifying structure ahead of change,” the anxiety is lower but the credibility burden is higher, because people will look for evidence that the structure actually changes.
  2. It affects legal exposure. In several jurisdictions, selection criteria, consultation duties and notice periods depend on how the business reason is characterised. A stated rationale that does not match later facts can be used against the employer.
  3. It sets expectations for the next round. A company that says it is preparing for automation implies future change. Employees, regulators and investors will watch for it.

Why cut when profits are rising?

The apparent paradox, rising forecasts and falling headcount, is the central story. A few explanations fit the available facts, although none is confirmed by the company.

  • Cost discipline in a slower-spending environment. Coverage noted that investors were concerned about slower spending growth. Cutting costs is a way to protect margins if revenue growth moderates.
  • Reversing a hiring surge. One report described the layoffs as coming after two years of hiring. If recent growth in headcount outpaced revenue, a correction is a predictable result.
  • Management-layer compression. Widening spans of control is a structural change that reduces the number of managers regardless of any technology plan. It is also a change many large companies have pursued since 2023.
  • Funding the technology investment. Savings from fewer roles can be redirected to systems, data work and automation, which in the near term may require both money and engineering talent.

Taken together, the signal is that strong results no longer protect headcount. Boards and executives treat workforce size as a continuous variable to be tuned rather than a cost to be cut only in downturns.

💡 Pro Tip: If you are asked to design a restructuring around “AI readiness”, write down the specific processes and decisions that will change, the metrics that will show it, and the dates. Vague rationales are hard to defend to employees, regulators and courts.

A sector-wide pattern

Amex was not alone that week. Barclays announced about 5,000 job cuts the day before, also referencing AI, and commentators described a wave of reductions across finance. The overlap is partly coincidence of earnings calendars: large banks and payment companies report in the same few weeks, and restructuring announcements often travel with results. But the consistent vocabulary, efficiency plus AI preparation, suggests that executives see a common structural pressure. Back-office, operations, customer-service and middle-management functions are the ones most often cited as candidates for change.

For HR, the sector pattern has practical consequences. Competitors in the same labour market are releasing experienced people at the same time, which temporarily improves hiring conditions for firms that are still recruiting. Conversely, companies that are retaining staff may find that their best people are being approached because the market narrative has made job security a worry.

What HR leaders should do now

1. Build a defensible rationale

Document the business reasons, the alternatives considered and the selection criteria before any announcement. Where consultation with employee representatives is required, begin early and keep records. Selection should rest on role, skills and objective criteria, with an adverse-impact review across protected groups.

2. Communicate in layers

Employees need three things: what is happening, how decisions are made, and what it means for them. Tell affected people first, individually and with support available. Then address remaining staff with honest information about whether further change is expected. Avoid treating “AI” as a catch-all explanation; specify what will change.

3. Protect the people who stay

Survivor effects, such as lower engagement, higher attrition among top performers and risk-averse behaviour, are well documented after large reductions. Managers with wider spans of control need training and realistic workloads, otherwise the efficiency gain disappears into burnout. Track voluntary attrition and engagement monthly for at least two quarters.

4. Invest in redeployment and skills

If automation is the long-term plan, identify the roles that will grow and the skills they need, and offer internal moves before external hiring. Outplacement support, extended benefits and skill grants reduce reputational damage and legal friction.

5. Prepare for scrutiny

Policymakers in several countries are paying closer attention to AI-linked redundancies, and employees’ representatives are asking for transparency. Expect questions about whether the cuts are really about technology, how affected workers were chosen, and what retraining is available.

Open questions

Several facts are still missing, and readers should treat early coverage with caution. Amex has not said which business units or countries are most affected, how long the programme will take, or how much it expects to save. The company’s quarterly filing and later investor presentations are likely to supply more detail. Until then, claims about the scale of automation behind the cuts are speculation.

Also unresolved is how investors will judge the move over time. The initial share-price decline suggests the market was not rewarded by the announcement, although a single day’s trading says little about the programme’s long-term merit. If the savings materialise and customer service does not deteriorate, the market may revise its view. If service suffers or further cuts follow, the narrative will turn.

Bottom line. American Express is cutting about 3,500 jobs while raising its profit outlook, and describing the goal as a leaner company prepared for AI. The honest reading is that it is a structural cost and organisation reset with AI as the stated direction, not yet proof that AI has replaced those jobs. HR teams should respond by tightening their own rationales, communicating specifically, and protecting the capability of the people who remain.

This article is for general information and is not employment-law advice. Figures are taken from press coverage of the company’s October 23, 2026 announcement and may change as filings are published.

Sources


Discover more from Kurums | Business Intelligence

Subscribe to get the latest posts sent to your email.

Discover more from Kurums | Business Intelligence

Subscribe now to keep reading and get access to the full archive.

Continue reading