Barclays Put Claude in Its Engineers' Hands: Regulated Banking Is All In on AI | inicIA

Direct answer

On October 1, 2026, Barclays expanded its collaboration with Anthropic to bring Claude across its global operations, with a public target: 50% of its developers on Claude Code before the end of 2026, and a majority of software engineers in 2027. Already today, more than 16,000 employees use its internal knowledge assistant (over one million searches since 2025), and Claude classifies and routes about 120,000 client emails a day in Global Markets. If a regulated universal bank is betting like this, the objection "AI isn't for regulated industries" just lost its best example.

The deployment numbers

The detail that matters most: the controls

Barclays didn't release AI without a safety net. The bank speaks of governance, security controls, and human oversight, and in regulatory documentation human review remains mandatory at every stage (under obligations like MiFID II). Group Co-Chief Operating Officer Craig Bright summed it up: they're moving toward AI as an increasingly agentic capability, embedded in how they build, test, secure, and operate technology. Translation: AI touches the bank's code, but with guardrails.

What to copy for your business (even if you're not a bank)

  1. Start with internal knowledge. Before touching customers, Barclays gave 16,000 employees a knowledge assistant. Your version: an assistant that answers questions about your processes, pricing, and policies. Low risk, high usage.
  2. Automate triage, not the decision. 120,000 emails classified and routed — the final call stays human. Same applies to your quotes, tickets, or WhatsApp messages: let AI sort and prioritize, let the human close.
  3. Set dated targets, like they did. "50% by year-end" is a measurable commitment. Define yours: which process, how many cases, by when. Without a target there's no adoption.

Related links

Frequently asked questions

What did Barclays announce with Anthropic?

On October 1, 2026, the expansion of their collaboration: bringing Claude and Claude Code to global operations, targeting 50% of developers on Claude Code by end of 2026 and a majority of engineers in 2027.

Are they already using AI at scale?

Yes: 16,000+ employees use the internal knowledge assistant, and Claude processes about 120,000 client emails a day in Global Markets. The engineering adoption targets are announced expectations; the real usage figures are the assistant's and the daily emails.

How do they handle regulatory risk?

With governance, security controls, and human oversight; in regulatory documentation human review is mandatory at every stage. The model is automating the heavy lifting while keeping the human on the decision.

Take the next step

If a regulated bank automates with controls, your business can too: internal assistant, message triage, and clear targets. Talk to inicIA on WhatsApp: https://wa.me/17866120930

By inicIA