🔍 Read the full analysis: UBS’s AI Agent Push: A View Into Everyday Bank Work on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
UBS reported that employees were running more than 12,700 AI personal agents as of its Oct. 28 third-quarter results, alongside 608 AI applications in production. The bank has not said what the agents can access, how they are reviewed or whether they can act without human approval; its CEO said UBS does not allow outside AI models to act for clients.
UBS said employees were running more than 12,700 AI personal agents to support daily work, according to the Swiss bank’s third-quarter results released Oct. 28. The disclosure puts a number on internal AI use at a major regulated financial institution, but UBS did not explain what the agents can access, what actions they can take or how their use is monitored.
The results release also reported 608 AI applications in production, an 81% increase from a year earlier, and more than 1,090 additional applications in development. UBS said more than 18,000 software developers were using AI tools in their development cycles, describing the productivity gains as measurable but giving no figure for them. The bank also reported nine large-scale AI initiatives and more than 30 million insights generated so far in 2026 by STAAT, a tool supporting its U.S. financial-adviser teams.
Reuters and PYMNTS reported the figure of more than 12,700 agents, citing UBS’s results. The release does not define “personal agent.” UBS had said in its July 29 second-quarter update that all employees had the ability to create their own AI agents, with the stated aim of helping staff spend more time on judgment, collaboration and client service.
UBS’s technology materials describe other employee-facing tools, including an in-house assistant called Red, rolled out to about 100,000 employees, and an AI-powered human-resources assistant for tasks such as checking policies and booking holidays through Microsoft Teams. Those examples provide some context for workplace AI at the bank, but UBS has not said whether they are included in the 12,700-agent count or how the agents in that figure are built.
Infographic 01 · What UBS reported
12,700 personal AI agents, inside one bank
The AI figures from UBS’s third-quarter results, and how fast they grew in three months.
From permission to 12,700 agents
UBS 2Q26 release (29 Jul 2026) vs. 3Q26 release (28 Oct 2026)
What the agent number is not
- Not 12,700 people. One employee can run several agents; UBS does not say how many people use them.
- Not a definition. UBS does not say what an agent can access or do on its own.
- Not a result. “Measurable productivity gains”, but no hours or costs published.
- Scale, for context: about 13 agents per 100 internal full-time staff (12,700 ÷ 97,593, own calculation).
“Our employees are running over 12,700 AI personal agents to support them in their daily activities.”UBS, third-quarter 2026 results release, 28 October 2026
Also in the release: 9 large-scale AI initiatives and more than 30 million insights from the STAAT tool for US financial advisor teams so far this year. UBS placed 6th of 50 banks in the 2026 Evident AI Index.
Sources: UBS 3Q26 release, 28 Oct 2026 · UBS 2Q26 release, 29 Jul 2026. Reported identically by Reuters, 28 Oct 2026 and PYMNTS (29 Oct 2026).
Powered by Thorsten Meyer AI
A Rare Measure of Workplace AI
The disclosure matters because it shows AI agents are not only a future proposal at UBS: the bank reports that employees are already running them at scale. The figure could help investors, workers and other companies gauge how quickly AI tools are being adopted inside a large organization, even though it does not show how often agents are used or what results they produce.
For a bank, the distinction between preparing work and taking action is central. A tool that drafts a summary for an employee presents a different risk from one that can send instructions, approve a transaction or change a client account. UBS has not disclosed where its personal agents fall on that spectrum. The number therefore signals adoption, but by itself does not establish the agents’ autonomy, reliability or effect on staffing and productivity.
The disclosure also points to a governance challenge for companies adopting employee-built tools: oversight must account for what systems and information each agent can reach, as well as whether its output is checked before use. Swiss regulator FINMA expects banks to maintain a complete inventory of AI applications and classify their risks. The agency has said that task has become harder as applications become widely accessible.
Popular right now
Vitamix 5200 Professional-Grade Blender, 64 oz. Container, Black
Bose QuietComfort True Wireless Noise-Cancelling Earbuds, Black
Dyson Car+Boat Cordless Handheld Vacuum
As an affiliate, we earn on qualifying purchases.
From Employee Access to Agent Counts
UBS’s July 29 results said all employees could create their own AI agents. In the Oct. 28 results, the bank reported more than 12,700 agents, 608 live AI applications and more than 1,090 in development. The release said the number of live applications was up 81% year over year. A comparison of the agent total with the July statement is not available: UBS did not publish a comparable agent count in that earlier update.
The third-quarter release placed the figures in a paragraph about investment in the bank’s business, alongside its financial and operational results. UBS also cited its sixth-place overall position in the 2026 Evident AI Index, which ranks 50 large banks. That ranking is separate from the bank’s own counts of agents and applications, and does not independently verify how those tools work.
UBS has also drawn a public boundary around outside AI systems acting for clients. On Oct. 27, CEO Sergio Ermotti said the bank was not allowing external models to enter the bank and act on clients’ behalf, Reuters reported. That statement concerns external models acting for clients; it does not, by itself, describe the permissions or controls applied to UBS employees’ personal agents.
“Employees “are running over 12,700 AI personal agents to support them in their daily activities.””
— UBS, in its third-quarter results released Oct. 28
What UBS Has Not Disclosed
UBS has not defined what counts as a personal agent or said whether the 12,700 figure represents agents created, active agents, or a different measure. The bank has not described which models and platforms they use, what internal data or systems they can access, how many employees use them, or how often their output is reviewed.
It is also unclear whether any agents can carry out tasks on their own or only prepare material for an employee to check. UBS has not said how many of its 608 live AI applications are client-facing, how it measures the reported developer productivity gains, or whether it has connected AI adoption to changes in staffing. The reported fall of about 1,500 full-time-equivalent staff in the quarter was attributed by UBS and Reuters to the Credit Suisse integration, not to AI.
As a result, the figures provide a snapshot of scale, not an account of the agents’ performance or risks. The bank’s statement about outside models acting for clients does not answer questions about the controls on internally used agents.
The Governance Questions Ahead
The next useful disclosures would clarify how UBS counts an agent, which tasks the tools can perform and what human approvals are required before an action is taken. Details about access controls, review procedures, incidents and measured productivity would help readers distinguish experimentation from routine, controlled deployment.
FINMA’s expectations for banks to maintain AI inventories and assign risk classes give a regulatory frame for those questions, though the supplied reporting does not identify a new UBS-specific regulatory action or deadline. UBS’s future results or technology disclosures may add detail; for now, the bank has reported the scale of employee use without publishing an operational breakdown.
For other businesses, the UBS figures offer a concrete prompt rather than a ready-made deployment model: determine who can create agents, what information they can use, what actions they can perform and who checks the results. Whether UBS’s agent count translates into measurable business or client outcomes remains unknown.
Key Questions
How many AI personal agents did UBS report?
UBS said in its Oct. 28, 2026, third-quarter results that employees were running more than 12,700 AI personal agents to support daily activities.
What can UBS’s personal agents do?
UBS has not defined the term or listed the agents’ capabilities. The bank has described employee-facing AI tools elsewhere, but it has not confirmed which of those tools are counted among the 12,700 agents.
Can the agents act on client accounts?
The results release does not say what actions employee agents can take. CEO Sergio Ermotti separately said UBS does not allow outside AI models to act for clients, such as managing accounts or moving deposits. That statement does not specify the permissions of internal employee agents.
Did UBS say AI caused staff reductions?
No. UBS’s internal headcount fell by about 1,500 full-time equivalents in the quarter, to 97,593. UBS and Reuters attributed the decline to the Credit Suisse integration, and UBS did not link it to AI.
How many AI applications does UBS have in production?
UBS reported 608 AI applications in production, up 81% from a year earlier, and more than 1,090 additional applications in development. The bank did not provide a breakdown of their uses in the results release.
Source: ThorstenMeyerAI.com
