Loading page…
The engineering push builds on staff information searches and client-email routing, bringing AI further into the bank’s existing technology and operations.
Listen to this story
Barclays’ expanded Anthropic collaboration puts Claude into staff knowledge retrieval, Global Markets email triage and software modernization, extending its use beyond a single internal assistant. The bank targets Claude Code adoption among 50% of developers by the end of 2026 and a majority of software engineers in 2027, but has not reported measured reductions in support time or engineering effort; the milestones track reach, not outcomes.
Barclays said its Colleague Knowledge Assistant, live since 2025, has been adopted by more than 16,000 colleagues and handled over one million searches; it is an internal staff tool, not a customer-facing chatbot.
The Global Markets platform processes approximately 120,000 client-enquiry emails daily, a volume figure that does not mean Claude resolves that many requests.
Barclays announced the expanded collaboration on October 1, 2026, and says governance, security controls and human oversight apply to its AI use cases.
Barclays is moving Claude further into the work behind its banking services, from finding answers for staff to modernizing older software. On October 1, 2026, the bank announced an expanded collaboration with Anthropic across its global operations. Its next adoption target is concrete: Barclays expects Claude Code to reach 50% of its developers by the end of 2026.
Barclays’ Colleague Knowledge Assistant has been live since 2025 and is powered by Claude. Its job is to help employees find information and answers more quickly while supporting the bank’s more than 20 million UK retail customers.
The bank says more than 16,000 colleagues have adopted the assistant, which has handled over one million searches. Those figures describe an internal tool used by staff, not a customer-facing chatbot. Barclays describes the benefit as faster information access and quicker customer support.
The Global Markets deployment handles a different task. Claude models classify incoming client-enquiry emails, enrich them with information and help determine their processing route. The system helps teams prioritize requests and check that they contain the information operations colleagues need to take action.
Barclays says the platform processes approximately 120,000 emails a day. It describes the benefit as less manual handling and easier identification of relevant information. The workload figure measures emails passing through the platform—not the number of client requests Claude resolves on its own.
After its year-end developer target, Barclays expects Claude Code adoption to reach a majority of software engineers in 2027.
The bank says governance, security controls and human oversight apply to its AI use cases. Group Co-Chief Operating Officer Anne Marie Darling frames the expansion as a way to reduce routine work and streamline decisions, allowing colleagues to focus on more complex problems.
The adoption targets set a timetable for wider use, not a measure of better software or service. Barclays has not quantified reductions in customer-support time or software-development effort in this announcement.
Loading discussion...
Join the conversation
Explain what would make widespread use count as genuine improvement.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.