The Ten-Minute Credit Line: How MYbank Is Rebuilding Small-Business Lending Around a Conversation

A restaurant owner in China wanted a larger credit line to finish building his second location. About ten minutes of conversation later, after human approval, he had a finalized financing plan offering a credit line of RMB 400,000, to be released in tranches as the new store progresses. The conversation was not with a loan officer. It was with Lark 2.0, an AI agent built by MYbank, a leading digital bank in China and an associate of Ant Group.
The case is now one of the working examples MYbank is showcasing at the Singapore FinTech Festival (SFF) 2026 — a step up from last year’s festival, where the bank presented proven applications of AI in credit risk control. The agent, launched in September 2026 at the Inclusion Conference on the Bund, opens credit, bill discounting, accounting and tax services to 42 million small and micro business owners. For a closer look at how the case unfolded, and what the bank rebuilt to make it possible, see the full account of MYbank’s Lark 2.0.
What the agent actually does is closer to casework than to a menu. Over a few rounds of dialogue, Lark 2.0 works through the owner’s business case and funding gap, assembles the supporting materials and submits them for review. The customer authorizes the agent to draw on operational information about the business. What comes back is a plan built around the need rather than a product chosen from a list.
Ma Xiaohang, Chief Information Officer of MYbank, frames the change in the shape of the answer. “Whether a bank serves its customers well depends not on willingness, but on capability,” he said. “Information about small businesses is scattered and volatile, and hard for banks to grasp fully and in real time, while a bank’s professional capabilities are spread across different departments and difficult to match and mobilize in time. The development of AI agents offers a new answer to these problems.” The result, in his words, is personalization: “Lark 2.0 can help create personalized solutions, taking inclusive finance from ‘standard answers’ calculated on existing information to answers ‘generated in real time’ based on needs.”
“Agent First” and the workbenches behind the chat window
The chat window is the visible part of a larger reorganization. Under an “Agent First” principle, MYbank sends information gathering, analysis and execution to AI first, while employees set standards, review results and make judgment calls at the key moments. Around that division of labor the bank has built eight AI workbenches, and it points to three results: its quantitative risk-control workbench cut the mishandling of legitimate business owners by about 40 percent; its human-review workbench lifted capacity for complex loan applications nearly 20-fold while keeping asset quality stable; and its product R&D workbench cut the time to a first verifiable version from 33 days to 2 days.
The industry is asking the same questions
The launch lands at a moment when financial institutions worldwide are working out how to move AI from pilots to scaled use. In June 2026, the World Economic Forum, working with Accenture, published a white paper, “The AI Playbook for Financial Services,” which pointed out that financial institutions’ AI adoption is moving from local experiments to organization-wide, scaled deployment — and that as AI capabilities become widespread, strengthening customer relationships and improving customer experience is becoming key to building differentiated advantages. MYbank’s arrangement, in which AI does the legwork while people keep the decision rights, sits squarely inside that debate.
How the bank got here
MYbank’s history explains part of the design. Founded in 2015 with a focus on serving SMEs and farmers, it was among the first group of pilot private commercial banks in China and the first bank in the country to deploy its core banking system entirely on the cloud, with no physical branches. It made its name on the “310” model: three minutes to apply for a collateral-free loan on a phone, less than a minute to approve, zero human interaction. The bank has also harnessed technologies such as AI and remote sensing to streamline credit services for farmers, eliminating lengthy processes and paper documentation.
What remains to be proven
The bank’s figures are its own, and the ten-minute case is a single case. Whether that speed holds for a first-time borrower with real but unaudited numbers is the question inclusive finance has always been judged on. What the launch does establish is direction: a leading digital bank now treats a conversation, not a form, as the front door of its lending business.


