Banks’ latest disclosures offer more detail on how AI is changing operations. FAB and BBVA reported faster processing and productivity gains, while Happen supported higher lending volumes with fewer operations employees. UBS provided evidence of adoption within wealth-adviser workflows.

Elsewhere, UK supervisors identified gaps in monitoring after deployment. Schwab expanded advisers’ access to AI tools, CIBC introduced a new core banking platform, and BNP Paribas extended its group AI platform. RBI and EBRD advanced programmes to strengthen their technology infrastructure.

Read more on the week’s key developments:

1. FAB cuts trade-document processing times by over 70% through AI

First Abu Dhabi Bank (FAB) disclosed on 29 October that AI-enabled workflows reduced trade-document processing turnaround times by over 70%. Employees had developed more than 6,000 AI agents as part of its enterprise-wide AI programme. The bank also reported growth exceeding 80% in new-to-bank onboarding volumes in business banking, supported partly by AI-enabled workflows.

FAB’s February AI update described an Innovation Hub overseeing reusable capabilities, more than 30 agentic use cases progressing across the bank and over 90% of structured data integrated into its enterprise platform. This common foundation supports deployment across business processes. In transaction banking, shorter document turnaround can help accommodate greater volumes and improve service delivery. The reported processing result gives FAB a concrete operational measure alongside the expansion of employee-developed agents.

2. BBVA reports AI productivity gains in software development and customer claims

BBVA reported on 29 October a 28% increase in software developer productivity from AI tools. Redesigned processes using its initial AI agents reduced the time to register customer claims by approximately 80% and to produce enterprise financial programmes by around 30%. The bank reported recurrent use of ChatGPT and Gemini by 75% of employees, with 2.4 hours saved per user each week. A contact-centre voice-interaction pilot achieved an approximately 90% successful resolution rate.

The claims result concerns registration, while the voice-interaction result remains pilot-based. BBVA’s Blue assistant also filters more than half of in-app conversations between clients and relationship managers, placing AI within an established service channel. Its Frame initiative provides a common approach to building, deploying and governing agents. Together, these developments show how BBVA is extending AI from individual employee tools into customer servicing and process redesign, supported by shared implementation and oversight arrangements.

3. UBS reports 608 live AI use cases as employee adoption expands

UBS reported on 28 October that it had 608 AI use cases in production, up 81% year on year, with more than 1,090 additional applications in development. Employees were running over 12,700 personal AI agents, while more than 18,000 software developers were using AI tools. The bank was advancing nine large-scale, end-to-end AI initiatives. Its STAAT advisory tool generated over 30 million insights year to date for distribution to financial adviser teams.

UBS’s AI overview describes a hub-and-spoke model in which its Chief AI Office co-ordinates governance and common capabilities, while business divisions own implementation and outcomes. STAAT applies this approach to wealth management, analysing internal and external signals, including property transactions and business funding activity, to help advisers prepare for client conversations. UBS says the tool serves more than 5,000 US advisers, with nearly 90% of adviser teams actively using it and an estimated 1,200 hours of preparation time saved weekly. These measures indicate how the tool is being incorporated into advisory work.

4. Bank of England and FCA find widespread AI-agent adoption but limited autonomy

The Bank of England and Financial Conduct Authority published their 2026 AI survey on 26 October, drawing on responses from 196 financial services firms, including 105 completing the full questionnaire. Overall, 86% reported using AI, compared with 75% in 2024, although respondent populations differed. Among full-survey respondents, 80% reported deployed AI agents, but only 2% reported agents with broad autonomy. Third-party implementations accounted for 53% of high-materiality AI systems, compared with 35% in 2024. The three largest cloud and computing providers represented 79% of reported relationships in that category.

Controls before deployment were more widely reported than ongoing monitoring. More than 90% reported human approval, pre-deployment review and vendor due diligence, against around half reporting ongoing monitoring, leakage testing or periodic review. Only 39% monitored customer-impact metrics, although not every application is customer-facing. The findings identify a gap between approving AI systems and overseeing their behaviour in operation, alongside substantial dependence on external technology providers.

5. Happen supports higher lending volumes with fewer operations employees

Happen, Inc., parent of Happen Bank, reported on 28 October that automation exceeded 90% for issued personal loans. In its earnings call, Chief Executive Scott Sanborn said the company supported 26% higher loan-origination volume with 20% fewer operations employees. Its Penny AI customer-service agent resolved 30% more calls than its previous system, while automated quality assurance covered all member-support calls. Happen also reported a 21% increase in code pull requests per technology employee and deployed an operations data analytics agent.

Sanborn estimated that processing the current volume at the previous year’s efficiency would have required more than 200 additional operations employees. That provides a measure of hiring avoided as lending expanded. Automated quality assurance also extends review across all support calls, giving the bank broader visibility into service problems and reasons for customer contact. Sanborn reported that AI use roughly doubled between July and September while expenditure fell, adding a technology-cost dimension to the operating gains.

6. Charles Schwab integrates AI tools into financial adviser workflows

Charles Schwab announced on 28 October that Claude for Financial Advisors was available through an authenticated connection to Schwab Advisor Center, allowing registered investment advisers to use Schwab account data for AI-assisted research, preparation and documentation. The company also introduced Wealth.com’s Ester tax and estate planning tool and announced a strategic partnership and investment in Nevis, which develops AI-powered operational workflows. Schwab Advisor Services supports more than 17,000 independent registered investment advisers.

The integrations are at different stages. Claude can retrieve authenticated Schwab account data, while Ester currently supports document analysis and guided planning. Authorised Schwab client-data access for Ester is expected in the first quarter of 2027. Alison Dooher, head of AI for Schwab Advisor Services, linked the approach to the firm’s existing integration infrastructure. Connecting tools to that infrastructure can reduce the manual transfer of information between adviser systems, with permissions governing access to client data.

7. CIBC adopts Thought Machine’s cloud-native core banking platform

Canadian Imperial Bank of Commerce (CIBC) announced on 29 October that it selected Thought Machine’s Vault Core platform for new banking services, becoming the first large Canadian bank to adopt the technology. The first application is The Simplii Account, a product from its direct banking division piloting this autumn, with digital savings folders and savings goals. CIBC said Simplii Financial developed the product in a matter of months. The agreement follows a GBP 30 million minority investment in Thought Machine earlier in 2026.

CIBC is introducing the core through a defined new product, giving its engineers operational experience with Vault Core within a limited initial scope. The bank said the platform would give its teams greater control over product development and delivery. Hratch Panossian, senior executive vice-president and head of personal and business banking, described the selection as part of changing how CIBC innovates. Its minority investment also extends the relationship beyond technology procurement, giving CIBC a financial interest in the platform provider.

8. BNP Paribas brings 100,000 people onto its group AI platform

BNP Paribas reported in its 28 October results presentation that 100,000 people had been onboarded to My Companion, its group AI platform. The bank described a multi-cloud, multi-model approach supported by shared platforms and reusable capabilities. Group AI strategic committees are chaired by its chief executive. Priorities include high-impact applications, cost control and cybersecurity, with AI identified as a component of its 2030 plan.

Its September Google Cloud partnership illustrates how this approach extends into business operations. Gemini is to be integrated into LLM@CIB, an assistant available to more than 65,000 employees, with agents planned for tasks such as preparing corporate credit memos. At Nickel, Gemini already supports 200 customer advisers in retrieving procedural guidance. Group Chief Information Officer Marc Camus said model and infrastructure choices would balance quality, security and cost. Some data is excluded from public clouds, while agents are to operate with restricted permissions and monitoring.

9. RBI inaugurates new data centre to strengthen financial-system resilience

The Reserve Bank of India (RBI) announced the inauguration of its new data centre and Cybersecurity Training Institute in Bhubaneswar on 25 October. Prime Minister Narendra Modi inaugurated the facilities. The data centre will house mission-critical systems, including Real Time Gross Settlement (RTGS) and National Electronic Funds Transfer (NEFT). The institute will develop cybersecurity, AI and emerging-technology expertise among RBI personnel and financial institution employees.

In an interview with DD India, Governor Sanjay Malhotra said RBI’s three existing data centres were in Maharashtra, while Bhubaneswar provides geographical separation in a different seismic zone. He described the facility as Tier IV-accredited and independent of third-party cloud providers. The additional location diversifies infrastructure supporting critical payment, settlement and regulatory applications, addressing concentration risk alongside the institute’s investment in cybersecurity skills.

10. EBRD appoints Cognizant for five-year technology programme

The European Bank for Reconstruction and Development (EBRD) selected Cognizant on 26 October under a five-year technology services framework supporting its Tech Blueprint transformation programme. The agreement covers cloud infrastructure, data, cybersecurity, software engineering, workplace services and managed IT operations. Cognizant plans to introduce Neuro IT Operations for proactive service management, Flowsource for AI-assisted software development and WorkNEXT for employee technology support. The framework combines outcome-led managed services and fixed-capacity delivery models.

Tech Blueprint concerns EBRD’s internal infrastructure and operations, distinct from its Digital Approach 2026–30, which guides investment, advisory and policy support for clients’ digital transformation. Simon Reeves, director of technology at EBRD, said Cognizant’s managed services expertise, AI-enabled delivery and results-driven approach would support Tech Blueprint and accelerate the bank’s modernisation and digitalisation. The agreement brings AI tools into a broader programme spanning software delivery, IT operations and employee support, with service outcomes forming part of the delivery framework.