Barclays to Cut 5,000 Jobs, Cites AI as a Factor

Barclays is set to eliminate approximately 5,000 jobs by 2026, representing nearly 5% of its workforce, citing artificial intelligence as a factor in this decision. The announcement coincided with the release of the third-quarter results on October 22, in which the British bank reported a pre-tax profit of £2.3 billion. According to CFO Anna Cross, around 60% of the job cuts will involve support functions, including software developers, as AI is making teams more productive.

The detail regarding developers is significant for those selling technology services. Software engineering used to be a protected area from the banks' cost-cutting programs. Now, it has made the list at Barclays.

The Figures That Barclays Already Released

The job reduction did not come as a surprise. On October 1, Anthropic announced an extension of its contract with Barclays, stating that the bank expects to have 50% of its developers using Claude Code by the end of 2026, reaching the majority of engineers by 2027. In the same statement, the bank described two uses already in production: an internal assistant adopted by over 16,000 employees, with more than one million inquiries, and a system that classifies around 120,000 emails daily in the Global Markets division.

"We are moving towards AI as an increasingly organic capability, embedded in how we build, test, secure, and operate technology," said Craig Bright, co-COO of the group, in the October announcement. Three weeks later, the same division appeared in the staffing reduction plan.

However, CEO C.S. Venkatakrishnan refrained from presenting the cuts as a direct effect of a chatbot. He stated that the reductions stem from years of investment in technology and focus on operational roles. AI, he remarked, is "the icing on the cake, not the cake itself": the real work is in the infrastructure that precedes it.

Goldman’s Counterpoint

On the same day, an analysis by economist Joseph Briggs from Goldman Sachs provided a strong argument against the notion that AI is already emptying offices. According to a report from Fortune on the analysis, Briggs estimates that up to 9% of workers could be displaced during the transition to AI, with another 3% effect due to slower hiring, over a period of about a decade. The key aspect here is timing: analyzing decades of data, he concluded that the disruption caused by automation is limited when the economy is healthy and accelerates during recessions.

Interpreted this way, Barclays’ case becomes less clear-cut. A bank posting quarterly profits of £2.3 billion cutting 5% of its workforce during a period of expansion contradicts the historical pattern described by Briggs or confirms that banks have always used cycles of technology investment to streamline operations, with or without AI. The CEO leans towards the second interpretation, while the CFO’s mention of developers leans towards the first.

There is one piece of data that undermines the mass replacement thesis: Barclays did not disclose how many of the 5,000 positions could be directly attributed to productivity gains from AI. Without that separation, any number regarding "jobs destroyed by AI" at the bank remains a third-party estimate.

Where the Effect Appears Outside the UK

Barclays is not alone in Europe. Dutch bank ABN AMRO announced plans in 2025 to eliminate around 5,200 jobs by 2028 and cited AI as one of the factors enabling them to maintain service levels with fewer employees. The timeline differs: ABN is spreading its reductions over three years, while Barclays aims to complete its cuts by 2026.

India is the second market to watch. Barclays maintains large technology and operations centers in Pune and Chennai, where much of the support and development functions now described as targets are located. If the pattern of "fewer developers due to AI" solidifies in captive centers, the effect would reach contracts with Indian integrators that share this work with banks, such as TCS and Infosys, both partners of Anthropic on projects for regulated sectors.

In the United States, where Briggs’ estimate represents about 15 million workers, the cyclical argument suggests that significant AI-driven layoffs would be postponed until the next slowdown. If he is correct, Barclays is a leading case, not the standard. If he is wrong, the British bank is the first of many to follow.

For consulting firms and integrators, the useful question is not whether AI will replace bank developers. It’s about who will capture the margins of the remaining work. If a client starts measuring productivity per engineer with a tool costing a fraction of a position, the hourly rate no longer becomes the standard, and development outsourcing contracts will be negotiated against the cost of a license.