🔍 Read the full analysis: UBS’s AI Agents: How Staff Use Them And What Remains Unsaid 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 in its Oct. 28 third-quarter results that employees run more than 12,700 AI personal agents, alongside 608 AI applications in production and more than 1,090 in development. The bank has not explained what the personal agents can access or do, how they are checked, or how many are used in client-facing work.
UBS said employees are running more than 12,700 AI personal agents in its third-quarter results published Oct. 28, putting a figure on staff use of agents at a major global bank. The disclosure also lists 608 AI applications in production and more than 1,090 in development, but does not explain what the agents can access, what actions they can take or how the bank reviews their use.
The figures appeared in a section of UBS’s 16-page results release about investment in its business. The bank reported nine large-scale AI initiatives, 608 live AI use cases—an 81% increase from a year earlier—and more than 1,090 additional applications in development. It also said more than 18,000 software developers use AI tools, describing the resulting productivity gains as measurable without quantifying them.
UBS reported that its US financial-adviser tool, STAAT, had generated more than 30 million insights so far in 2026. These metrics describe different kinds of activity: agents used by employees, applications in production or development, developer tools and outputs from a specific advisory tool. The release does not define how UBS counts an agent or an application, so the figures should not be treated as directly comparable measures of adoption or impact.
The bank had said in its July 29 second-quarter release that all UBS employees had the ability to create their own AI agents. That statement described access to a capability, not the number of employees using it. The October figure counts agents, but UBS does not say how many staff members run them, whether agents are shared, or how often they are used.
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
The Oversight Questions Behind the Count
The reported total matters because it shows agent use has moved beyond a general promise of future adoption at UBS: the bank says employees are already running thousands of these tools. But a large count alone does not establish how much work agents perform, whether they improve results or what risks they create. UBS supplied no quantified productivity result for the agents themselves.
For staff, the distinction between an assistant that prepares a draft and one that can take an action is consequential. A tool that summarizes material for a person to review has a different risk profile from one that can send a message, change a record or approve a transaction. UBS has not said which capabilities its personal agents have, so the public disclosure cannot show where they sit on that spectrum.
The disclosure also gives other companies a practical governance question: how to track tools employees can create themselves. UBS’s figures are a measure of reported activity, not proof that all agents are independently acting or that they have replaced jobs. The bank reported that staff numbers fell by about 1,500 full-time equivalents during the quarter, to 97,593, and attributed the decrease to the Credit Suisse integration—not to AI. No causal link between the staff reduction and AI was reported.
Products Worth Considering
Samsung 49-inch Odyssey OLED G9 (G91SD) Dual QHD Curved Gaming Monitor
LG UltraGear 34GX90SA-W 34-inch WQHD Curved OLED Gaming Monitor
Samsung 49-Inch Odyssey G93SC QD-OLED Curved Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Employee Access to Reported Use
The July statement that all employees could create agents preceded the October tally by roughly three months. UBS’s reported number of live AI use cases rose from 560 in the second quarter to 608 in the third; applications in development increased from more than 920 to more than 1,090. The bank did not provide a quarter-on-quarter agent count in the source material, so it is not possible to determine how quickly the 12,700 figure grew.
UBS’s technology materials describe an internal assistant called Red, rolled out to about 100,000 employees, and an AI-supported human-resources assistant that lets staff book holidays and look up policies in Microsoft Teams. These examples provide some context for internal AI use, but UBS has not said that either tool accounts for the personal-agent total or defined how those systems relate to the agents announced in its results.
There is a separate, more explicit boundary around outside AI systems acting for customers. At an ETH Zurich event on Oct. 27, CEO Sergio Ermotti said UBS was not allowing external models to enter the bank and act on clients’ behalf, such as managing accounts or moving deposits, Reuters reported. That statement concerns outside models acting for clients; it does not explain the permissions or controls applied to employee-run agents.
“Employees “are running over 12,700 AI personal agents to support them in their daily activities.””
— UBS, in its third-quarter results release
What UBS Has Not Disclosed
The results release does not define “AI personal agent” or state which models and software power the tools. It does not say what internal systems or data they can access, whether they can take actions independently, or whether a human must approve each output or action. Without those details, it is not possible to tell how many are drafting and organizing work versus doing tasks with direct consequences.
UBS also has not disclosed how agents are reviewed, what safeguards apply, how errors are logged, how many employees use them or how many of the 608 production applications involve clients. The reported productivity gains among developers are not quantified, and the bank has not provided a measured productivity outcome for the personal agents. Those are gaps in the public information, not evidence that controls or measurements do not exist.
The source material says Swiss regulator FINMA expects banks to keep a complete inventory of AI systems with risk classifications and has warned that inventories have become harder as applications become accessible to more users. The precise controls UBS applies to its reported agents—and how they map to regulatory expectations—are not described in the results release.
Further Detail on Controls and Use
The next useful information would be a clearer account from UBS of what it counts as a personal agent, what permissions agents receive and how the bank monitors them. Further disclosure on approval requirements, human review, incident handling and the split between internal and client-facing uses would help readers assess the scale beyond the headline count.
UBS’s reported metrics can be compared with later results as the bank updates its figures, but the current release does not set out a timetable for additional details. Until it does, 12,700 is a bank-reported count of agents, not a measure of how many staff use them, how much time they save or how much autonomy they have.
Key Questions
What did UBS report about AI agents?
UBS said in its Oct. 28, 2026, third-quarter results that employees run more than 12,700 AI personal agents to support daily activities.
What can the UBS agents do?
The results release does not define their capabilities. UBS has not said what systems they can access or whether they can take actions independently, so specific tasks should not be assumed from the reported count.
Does the 12,700 figure mean that many employees use agents?
No. UBS reported the number of agents, not the number of employees using them. The bank has not disclosed how many staff members run agents or whether some agents are shared.
Are the agents allowed to manage UBS clients’ accounts?
UBS has not described the permissions of its employee-run agents. CEO Sergio Ermotti separately said the bank does not allow outside AI models to act for clients, such as managing accounts or moving deposits, Reuters reported.
Did UBS link AI use to its lower staff count?
No. UBS’s internal workforce 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; the reported material does not link it to AI.
Source: ThorstenMeyerAI.com
