Artificial intelligence is rapidly evolving from simple banking chatbots into systems capable of acting on behalf of customers. With the launch of its new AI-powered assistant, Starling Bank is introducing a more autonomous model of digital banking where users can manage finances, organise savings, and complete tasks through natural language commands.
Digital banking in the United Kingdom has entered a new era of automation with the launch of the Starling Assistant. This tool represents the first “agentic” AI financial assistant in the British market, moving beyond traditional chatbots to a system capable of executing complex banking tasks on behalf of the user. Built using Google Gemini on the Google Cloud platform, the assistant allows personal current account holders to manage their day-to-day finances, organise bills, and set savings goals through simple voice or natural language prompts within the app.
The assistant integrates eight years of internal AI development, including the specialised Spending Intelligence and Scam Intelligence tools the bank released last year. By adopting an agentic model, the bank enables the AI to take proactive steps rather than just providing information. For example, a customer can instruct the assistant to set up a dedicated “Space” for a holiday and calculate the necessary monthly transfers to reach a specific target by July. The AI then automatically handles the technical setup of these transfers, reducing the manual effort required for effective money management.
Harriet Rees, the Group Chief Information Officer at Starling, emphasised that this technology is designed to help people be “good with money” by fostering better financial habits. Beyond simple transactions, the assistant provides deep personalised insights, such as analysing direct debit frequencies or transaction histories with specific payees. It also features a unique engagement layer where users can take quizzes about their own spending patterns to increase their financial awareness. For customers with accessibility needs, the assistant can guide them through setting up sign language services or implementing gambling blocks without requiring a human agent.
Security and privacy remain central to the rollout of this autonomous functionality. In line with the bank’s data ethics commitment, the assistant is an opt-in feature, and all customer data remains securely within the bank’s dedicated cloud environment. Crucially, this data is not used for training external models. Raman Bhatia, the Group CEO, noted that agentic AI is the next logical step for the industry, offering transformative potential to protect customers from scams and help them understand their financial health in real time.
The bank plans to expand these capabilities to business and joint account holders in the near future. For business users, the assistant will eventually be able to analyse spending by tax year and provide detailed payment history reports. This roadmap suggests a shift toward a “single interface” model, where all banking interactions—from ordering a new card to reporting a suspected scam—are handled through a unified, intelligent conversational layer. As the first UK bank to deploy this level of autonomy, the institution is setting a high competitive bar for the “Big Four” and other neobanks alike. By combining proactive task execution with sensitive support for vulnerable customers, the platform demonstrates how AI can deliver both convenience and essential financial inclusion. The move positions the bank at the forefront of the global trend toward “Command Line Commerce,” where the boundaries between banking software and personal financial advice continue to blur.
What this means for the industry
- Agentic AI marks the next phase of digital banking: Moving beyond chatbots, AI assistants are beginning to execute tasks such as setting up savings plans or managing account actions automatically.
- Conversational banking is becoming the primary interface: Natural language prompts and voice interactions may soon replace traditional app navigation for many routine banking tasks.
- Personalised financial insights are expanding: AI systems analysing transaction behaviour can deliver tailored recommendations to help customers improve money management.
- Security and trust remain critical: Opt-in models and strict data governance will be essential as banks deploy more autonomous AI capabilities.
- AI could transform financial inclusion: Features that support accessibility, scam protection, and behavioural awareness demonstrate how intelligent assistants can support vulnerable customers.
- Competitive pressure on traditional banks: As neobanks introduce advanced AI-driven services, larger institutions may be forced to accelerate their own AI strategies to keep pace.
Photo by BoliviaInteligente on Unsplash
