LHV Bank, a digital-only lender, has deployed an agentic AI system from Gradient Labs to handle customer email support in the UK, providing round-the-clock responses to routine banking queries.
The launch follows a proof of concept announced by the two companies earlier this year. The AI agent can now resolve everyday customer requests from start to finish, including reporting a missing card, clarifying payments, answering questions about transfers and providing updates on existing enquiries. Cases involving sensitive issues, complex requests or actions outside its approved procedures are escalated to human support staff.
Previously, LHV handled customer emails through a shared queue, meaning straightforward enquiries and more complicated cases entered the same process. The new system is designed to resolve routine requests immediately, allowing employees to focus on cases that require human judgement.
The agent operates within defined customer journeys and approved procedures, tools and controls. LHV’s teams design the workflows, establish the limits of the system and determine which actions it can perform. Every update is tested and approved by the bank before release, while teams continue to monitor service quality and review escalated cases.
To support its use in a regulated banking environment, the system combines large language models with deterministic guardrails. These controls restrict the data and tools available to the agent, define when it must stop or escalate a request, and determine how complaints and vulnerable customers are handled.
LHV said customer information is not used to train the models or build the agent. Instead, the system accesses relevant customer data only during live interactions.
The bank has also adopted Collaborate, a component of the Gradient Labs platform that applies software engineering practices to customer-facing AI. Teams across operations, product, engineering, risk and compliance can prepare changes in controlled environments, test them against representative customer scenarios and review them before approving a release. Changes can also be reversed, while records capture what was modified, who authorised it and when.
LHV plans to expand the agent’s responsibilities to cover additional servicing procedures within its UK email channel. Longer-term plans include exploring live chat and voice support, as well as back-office and financial crime processes such as onboarding, know-your-customer (KYC) checks and anti-money laundering (AML). The companies will also consider extending the technology to LHV’s Estonian operations.
LHV Bank interim CEO Kris Brewster said, “As a digital-only bank, we don’t have to accept the way customer service has always been done. Others in the industry are using AI co-pilots that simply draft replies for staff to approve. This might speed up individual steps, but customer queries still sit in the same queue, and when the AI is usually right, people can start approving suggestions without checking them closely. We wanted to go further and give customers an answer straight away, while keeping our people firmly in control of how the service works and stepping in whenever a case needs them. Getting that balance right is what responsible innovation looks like to us.”
Gradient Labs CEO Dimitri Masin added, “The future of AI in financial services is not a faster human approving AI suggestions at every step. It is trusted agents taking responsibility for defined work, supported by deterministic controls, clear escalation points and human judgement where it matters. LHV is showing what that can look like in practice.”
Copyright © 2026 FinTech Global
