Banks are increasingly exploring agentic artificial intelligence as a way to automate complex operational tasks while maintaining strict regulatory oversight. Unlike traditional chatbots, these systems are designed to reason through workflows and assist employees with decision-making, making them particularly suited to high-volume functions such as customer service and case management.
LHV Bank has officially announced a new Proof of Concept in partnership with Gradient Labs to explore the sophisticated use of agentic artificial intelligence. This initiative primarily supports retail customer service teams, with a specific emphasis on managing email-based communications. The collaboration marks a significant step for the London-based bank as it seeks to leverage advanced automation to handle customer inquiries more efficiently and consistently. Unlike traditional chatbots, agentic AI systems are designed to perform complex tasks and make reasoned decisions within a set framework. The initiative will explore opportunities to improve response quality and reduce turnaround times while ensuring human oversight remains central to the process.
LHV Bank and Gradient Labs intend to test selected use cases in a strictly controlled environment to maintain high safety standards. There is a strong emphasis on explainability and auditability throughout this trial period. This ensures that every action taken by the AI can be traced back to its underlying logic for regulatory and internal review purposes. Clear accountability remains a priority to ensure the technology aligns with the bank’s existing governance and risk requirements. Kris Brewster, the Interim CEO of LHV Bank, stated that customer support is a critical touchpoint for retail banking. He noted that the potential use of agentic AI represents an exciting opportunity to enhance the bank’s service offering while keeping human judgment and customer protection firmly in place.
The bank has a documented history of prioritising modern technology to improve its internal and external operations. In March of last year, the institution signed a deal with OpenAI to license ChatGPT Enterprise. This agreement provided all employees with unlimited access to generative tools to foster innovation across departments. More recently, the bank began using specialised technology from Luminance to streamline its legal contract and document workflows. These previous steps indicate a broader strategy to embed intelligent systems into the bank’s core functions. By moving toward agentic AI, LHV Bank is looking beyond simple text generation to systems that can actively assist with problem-solving and administrative execution within the customer service department.
The success of this Proof of Concept could lead to a wider rollout of agentic systems across other communication channels within the bank. As retail banking becomes increasingly digital, the ability to provide rapid and accurate support is a key differentiator for modern financial institutions. The focus on email-based inquiries allows the bank to refine the AI’s tone and accuracy in a format that permits thorough review before a message reaches a customer. This milestone confirms the trend of established banks seeking partnerships with specialised fintech firms to solve specific operational challenges. By combining the institutional credibility of LHV Bank with the technical agility of Gradient Labs, the project aims to set a new standard for how AI can support human teams in a highly regulated environment.
Key takeaways
- LHV Bank has partnered with Gradient Labs to test agentic AI for customer support operations
- The Proof of Concept focuses on automating and improving email-based customer service responses
- Agentic AI systems are designed to perform complex tasks and assist with decision-making rather than simple chatbot interactions
- The pilot emphasises explainability, auditability, and human oversight to meet regulatory requirements
- Builds on LHV Bank’s broader AI strategy, including ChatGPT Enterprise deployment and Luminance legal automation tools
- The initiative aims to reduce response times while improving consistency and service quality
- Successful trials could lead to expanded AI deployment across additional banking communication channels
