American Express is cutting 5% of its workforce to 'prepare for AI'

American Express is eliminating 5% of its global workforce and taking a $336 million restructuring charge. The company says the change is not only about cutting costs. It wants a simpler, more agile organization that can adopt artificial intelligence more effectively. The goal matters because poorly structured work can make automation slower and less useful.

Chief Financial Officer Christophe Le Caillec said Amex is reducing organizational layers and increasing spans of control. In simple terms, fewer management levels and broader responsibilities should make processes easier to redesign. That creates a cleaner foundation for automating repetitive work and applying AI to internal operations.

The company had almost 77,000 employees at the end of 2025, so the reduction could affect roughly 3,850 people. Amex says the changes will support innovation over the coming years. Its leaders also want AI to personalize customer experiences, while keeping outside agents limited to approved system connections.

A 5% workforce reduction at nearly 77,000 employees would affect roughly 3,850 people. The calculation is simple: 77,000 multiplied by 0.05 equals 3,850. Because the article uses “almost 77,000,” the actual number could be slightly different.

This estimate represents the potential scale of American Express’s restructuring. The company announced a global workforce reduction rather than a cut limited to one country or department. It is also taking a $336 million restructuring charge, showing that the change involves significant near-term costs.

The reduction is linked to a broader effort to become leaner and prepare for AI. Amex says it will reduce organizational layers and increase spans of control before automating more work. The article does not provide a final number of affected employees or a breakdown by role, so 3,850 remains an approximate calculation.

When AI automates repetitive internal work, companies may need fewer people for tasks that follow predictable rules. They may also reorganize jobs rather than simply remove them. This can lower routine workload, but it can create disruption for employees whose roles are changed or eliminated.

American Express gives a concrete example. It is cutting 5% of its workforce, reducing organizational layers, and increasing spans of control. The stated mechanism is to avoid automating badly designed processes. By simplifying the structure first, Amex believes AI can handle more work with fewer procedural barriers.

The operational effect is intended to be greater efficiency and agility. Amex’s CEO said the company wants to accelerate innovation, while its finance chief said automation could expand in coming quarters and years. The broader financial-services sector is pursuing similar technology-driven cost initiatives, including Barclays’ planned 5,000-job reduction.

Artificial intelligence is technology that analyzes information and performs tasks that often require human judgment, such as recognizing patterns, making predictions, or producing recommendations. This definition comes from general knowledge; the article focuses on how Amex plans to apply AI rather than defining it.

For customers, Amex’s CEO emphasized personalizing experiences and offers for cardholders. AI could examine permitted customer and transaction information to identify relevant recommendations. Inside the company, it could automate repetitive processes after those processes are simplified. Outside agents could also interact with Amex systems, but only through approved connections.

The company is not allowing uncontrolled access. Stephen Squeri said assistants will not “run rogue” through Amex’s systems. He said agents will be allowed to use connections approved through API or MCP channels. The approach combines convenience and automation with controls intended to protect customers and merchants.

Banks and card companies can pursue AI even during profitable periods because technology changes how efficiently they operate. Strong consumer spending supports revenue, but it does not guarantee that existing processes are fast, simple, or ready for future competition. Companies may invest or restructure before pressure becomes severe.

American Express links its move directly to efficiency and innovation. It is reducing 5% of its workforce, cutting layers, and increasing spans of control. Barclays has also confirmed plans to cut 5,000 jobs, about 5% of its staff, in a technology-driven cost initiative. These examples show how AI is becoming part of operating strategy.

The timing is notable because the sector has robust profits, growing consumer spending, and resilient credit. Amex still topped adjusted earnings expectations and raised its full-year guidance. Yet executives said the company needed to become more agile and efficient to win over the next few years. AI is therefore presented as preparation, not merely crisis response.

An API is a defined interface that lets one software system request data or actions from another under set rules. MCP generally refers to the Model Context Protocol, a structured way for AI applications to connect with tools and information. These definitions use established technology knowledge; the article identifies both as secure connection channels but does not expand their names.

For example, an approved AI agent might request a permitted account service through an API rather than freely entering Amex’s systems. The connection can limit what the agent sees and does. MCP can provide another controlled route for an agent to work with approved tools or information. The exact permissions depend on Amex’s design.

Stephen Squeri said Amex will not let assistants “run rogue.” Requiring approved API or MCP connections helps the company control access, protect customers and merchants, and monitor automated activity. It also lets Amex pursue useful AI features without treating every outside agent as fully trusted.

A credit-card company generally earns revenue from customer spending and fees connected with card use, merchant relationships, lending, and related services. This explanation uses established industry knowledge; the article specifically mentions growing consumer spending, merchant protection, fees indirectly through partnerships, revenue, and operating efficiency. Profit is what remains after expenses.

More spending can support revenue, while strong merchant relationships can help sustain card use and business activity. Costs move in the opposite direction. Employee expenses, restructuring charges, technology investments, and other operating costs can reduce profit. Amex’s $336 million restructuring charge is one immediate cost tied to its reorganization.

The company reported total revenue of $20.06 billion, up 9%, though slightly below the forecast. Executives linked some spending weakness to exits from small-business co-brand partnerships. Adjusted earnings per share beat expectations, and full-year guidance rose, helped partly by proceeds from selling Amex’s Global Business Travel Group stake.