Innovation and AI at UBS

Scaling high-quality research content with AI

AI is transforming UBS research workflows – boosting productivity, increasing engagement, and enabling analysts to deliver high-quality insights faster and at scale.

By embedding AI directly into the tools analysts already use, teams can now draft, refine, translate and distribute research content more efficiently and consistently.

The award-winning* initiative began with simple yet high-impact features such as AI-assisted titles and email hooks. These tools have now been used more than 15,000 times, and have increased email open rates by 16 percent.

Work then expanded into developing a broader analyst toolkit that now supports nearly 1,000 Global Research professionals. Our CoAuthor tool, used more than 30,000 times, helps strengthen clarity, consistency and overall content quality.

And we didn’t stop there. AI has also reshaped video production, moving from standard studio production to a more scalable model, using AI-generated analyst avatars and automated script generation. This enables us to produce significantly more video content at scale, helping meet strong and growing client demand for insights delivered in video format, while continuing to drive high engagement.

*Our work in this space has been recognized with a 2026 Celent Model Sell Side Award for GenAI Automation.

At UBS Global Wealth Management, we are harnessing artificial intelligence to help advisors deliver greater value to clients – more efficiently, more proactively, and with greater confidence. By transforming vast amounts of data into timely, actionable intelligence, AI enables advisors to focus less on information gathering and more on what matters most: building meaningful client relationships and delivering personalized advice.

Equipping advisors with timely client intelligence

One of our flagship AI solutions, STAAT Insights^*, provides more than 5,000 UBS financial advisors in the US with highly personalized intelligence at the moments it matters most. The platform analyzes thousands of internal and external signals to identify relevant client opportunities – from significant life events and property transactions to late – stage business funding activity.

Advisors receive concise, AI-generated briefings directly in their inboxes ahead of client meetings and can access insights on demand throughout the day.

Delivering global intelligence with local relevance

In Hong Kong and Singapore, our Conversational Insights* solution brings similar capabilities to more than 200 client advisors. By consolidating client, portfolio, and activity data into clear, actionable intelligence, the platform helps advisors quickly identify meaningful opportunities and service needs.

Whether highlighting upcoming maturities, cash-flow events, client milestones, or portfolio changes, Conversational Insights enables advisors to engage clients at the right time and with greater context. The result is a more proactive approach to client servicing and more personalized client interactions. Since launch, advisors in these markets have reported meaningful reductions in preparation time, while gaining greater confidence in delivering tailored, timely and relevant client conversations. 

*Our leadership in AI-enabled wealth management has been recognized globally, with 2026 Celent Model Wealth Manager Award for Data, Analytics and AI, Best Wealth Management Firm for Use of AI in the US, Financial Times PWM Wealth Tech Awards, Best International Private Bank for AI Adoption and Digital Innovation, Asian Private Banker Awards for Distinction 2025, Asia’s best for digital solutions 2026 | Euromoney Private Banking Awards 2026

Making enterprise knowledge instantly accessible

Red, UBS’s in-house AI assistant, helps employees access trusted internal knowledge quickly and complete knowledge-intensive tasks more efficiently. It brings together curated UBS content and generative AI capabilities in one place, making it easier to find information, draft content and move work forward.

By combining official UBS sources – including CIO research, process and product information, and factsheets – with capabilities such as summarization, drafting and translation, Red simplifies how employees find information and carry out everyday tasks. This reduces time spent searching across fragmented sources and supports faster, more consistent execution.

The solution has been rolled out to around 100,000 employees and has supported more than 25 million queries since launch, demonstrating broad adoption across the firm.

Based on user feedback, Red saves employees an average of 80 minutes per week, freeing up time for more meaningful client interactions and deeper engagement with clients' individual needs.

Faster, simpler and more personalized HR services for employees

AI is helping UBS reimagine the employee journey across Human Resources – from everyday support and onboarding to recruiting and performance management. By embedding generative AI and automation into the tools and processes employees and managers already use, we are making HR services faster, simpler and easier to access.

AskHR, our GenAI-powered HR assistant – Around 100,000 employees can now book vacation, review leave plans and access policies, as well as get guidance 24/7directly through Microsoft Teams. With close to an 80 percent in-app resolution rate, AskHR is improving the employee experience and reducing operational workload.

Candidate experience (AI-enabled onboarding) – We are automating key steps in the onboarding process to make the experience faster for candidates and HR teams alike. In multiple countries, automated contract generation for many hires has reduced time to contract from 24 hours to approximately 20 minutes, with built-in consistency, compliance and human oversight.

Performance management (GenAI-assisted evaluation summaries) – Generative AI helps line managers turn large volumes of structured and unstructured feedback into clear, high-quality year-end evaluation summary drafts. While managers remain fully accountable for the final assessment, the solution reduces manual effort and improves consistency, quality and compliance. It is already saving over 32,000 hours of line manager time each year.

Using AI to simplify complex work across Sourcing, Communications, Legal, Finance and Compliance

Across global functions, UBS is applying artificial intelligence, generative AI and automation to accelerate and simplify complex, knowledge-intensive work. From financial data sourcing and procurement automation to media intelligence and legal workflows, AI is helping teams reduce manual effort, improve consistency and deliver faster outcomes at scale.

Speeding up sourcing – Our Group Real Estate and Supply Chain (GRESC) team introduced a GenAI-enabled solution, ‘Front Door’, that guides employees through creating sourcing requests and assessing risk applicability. By simplifying inputs and automating guidance, the required manual entry fields have been reduced by almost 40 percent, and processing sourcing initiatives is now up to 60 percent faster.

Embedding AI into our Legal workflows – The purpose-designed platform ‘Harvey’ within Group Legal enables high-quality drafting, summarization, and review of complex legal content, powered by models trained specifically on legal language and structures. The tool has direct access to jurisdictional knowledge, enables large-scale document processing via a secure vault, and automates sophisticated workflows through reusable prompt chains. It seamlessly integrates with day-to-day applications, embedding AI directly into the core legal workflow. The increased efficiency delivers direct cost savings of at least USD 1 million. It also enables the insourcing of external legal work, generating an additional USD 1 million in savings in 2026 alone, while significantly improving decision-making and risk anticipation. Through our Legal translation tool, teams have direct access to professional-grade, context-aware translations across core languages, reducing reliance on external services and accelerating cross-border collaboration – saving an estimated USD 1 million annually through efficiency gains and reduced spend.

AI‑powered live event translation – so clients and audiences can follow UBS events along in their own language – enhancing accessibility and enabling real‑time multilingual experiences for a truly global audience, used at flagship events such as the UBS Greater China Conference 2026 and the UBS Asian Investment Conference 2026.

Accelerating our Finance transformation – The ledger is the backbone of Finance, essential to our financial and regulatory reporting. Modernizing our ledger requires sourcing and transforming large volumes of data from hundreds of business systems and thousands of data feeds – a significant task. But, through applying AI, we have been able to accelerate this work by around 50 percent.

MyPolicies AI assistant transforms how employees access governance information by replacing manual document searches with a personalized, AI-powered assistant that delivers natural-language answers with cited sources tailored to user roles and locations. Since its launch in August 2024, the solution has supported more than 288,000 conversations (around 418,000 questions), helping employees find accurate guidance faster while reducing time spent navigating complex policy documents. With access to over 1,100 policies and guidelines and more than 110,000 users across the bank, the tool demonstrates strong and sustained adoption at scale. Based on an average saving of 10 minutes per conversation, the solution has delivered approximately 48,000 hours saved since launch, while improving consistency in how employees access policy information.

Bringing AI into everyday work

Within technology, GitHub Copilot has become a cornerstone of the AI-augmented engineering strategy.

The AI-powered coding assistant is available across approved development environments and supports engineers throughout the software development lifecycle. Common use cases include code generation, refactoring, debugging, documentation, testing and application modernization.

The rollout has been supported by structured onboarding, training resources and active communities of practice designed to help engineers maximize the value of AI-assisted development. As a result, AI is becoming an increasingly embedded part of day-to-day engineering work across UBS.

Adoption has continued to grow, with approximately 18,000 engineers enabled on GitHub Copilot across the firm. Usage data also points to strong engagement, with around 89 percent of licensed users actively coding and consuming available tokens, exceeding both current targets and expectations. This level of participation demonstrates the growing confidence engineers have in AI-assisted development and the value they are seeing from these tools in their daily work. 

CORISC XAI is UBS’s AI solution for transactions screening. Designed for speed, transparency, and trust, it supports screening in near real time, enabling a 24/7 operating model and helping meet the growing demand for instant payments.

Across the Compliance Risk Screening ecosystem, more than 10 million alerts are generated each year. CORISC XAI helps manage this volume more efficiently while maintaining robust control standards. A significant proportion of alerts generated by traditional screening systems are false positives. CORISC XAI helps reduce this burden by analyzing and releasing low-risk alerts, allowing analysts to focus their attention on higher-risk cases. Our robust governance, monitoring and oversight framework allows us to remain in full control and have the right guardrails at all times. As a result, around 20 percent of alerts are handled automatically, resulting in productivity gains of approximately 30 percent.

By providing clear, audit-friendly explanations for every decision, CORISC XAI enables analysts to understand the rationale behind outcomes. This improves transparency, strengthens confidence in decision-making, and supports evolving sanctions and regulatory expectations. The solution is scalable and uses fully anonymized information, creating opportunities for broader adoption across the organization.