Barclays is cutting 5,000 jobs by the end of 2026, including software developers, as AI begins to reshape the economics of running one of Britain's biggest banks. For tech startups, the implications are that the effects of AI are going mainstream.
The cuts announced amount to roughly 5% of Barclays' global workforce. Around 1,000 positions have already disappeared, with another 4,000 set to go before the end of the year. Some 60% of the reductions involve support functions, including developers whose productivity has increased through AI.
Now, CEO CS Venkatakrishnan has been careful not to attribute all the savings to AI. He says the bank has spent years modernising its infrastructure, simplifying processes and consolidating systems.
AI, he said, is the "icing on the cake, it's not the cake".
But the icing is starting to look rather substantial.
In fact, the icing is borderline taking over the cake.
Earlier this month, Barclays expanded its partnership with Anthropic, announcing plans to deploy Claude Code to 50% of its developers by the end of 2026, rising to a majority of its software engineers in 2027. How many of those developers will be cut in subsequent months, I wonder.
Anthropic's technology is already said to be helping more than 16,000 Barclays employees retrieve information, while AI systems process approximately 120,000 emails daily in its Global Markets business.
Summing all this up, the direction of travel is clear. The bank wants more output from fewer people.
As Venkatakrishnan acknowledged during an investor conference in September, AI will ultimately "change the size and the shape of our workforce".
This is happening despite Barclays reporting a 12% rise in third-quarter pre-tax profits to £2.3 billion. So the cuts are about improving productivity and returns. It’s not as if they were short of cash to maintain the existing workforce.
For tech startups, there are some important implications.
The first is obvious, due to the cutting edge nature of startups. We already know software startups no longer need the engineering teams they once did.
Secondly, and more importantly, the moves by Barclays illustrate a growing opportunity for AI startups selling into large enterprises.
“Game on!” I can hear a few entrepreneurs cry.
The bank's experience suggests the real commercial value of AI is in integrating with legacy systems, automating workflows and reducing operating costs. This is all obvious.
But there’s a third implication. Could this contraction of large enterprises create a tougher market? Startups cannot sell generic AI productivity tools, which will be eaten up by foundational model providers such as Anthropic, which are increasingly building direct relationships with major corporate customers. They must go niche and vertical.
But Barclays’ announcement is proof of a major corporation explicitly connecting AI-driven productivity with a smaller workforce.
That means plenty of other large companies are going to be looking for the same kinds of savings and new, startup partners.