Blend Labs, Inc. (NYSE: BLND), a digital origination platform for banks, credit unions, and mortgage lenders, has announced the launch of Autopilot MCP. This server, built on the Model Context Protocol, provides authorized AI agents with secure, programmatic access to the entire Blend platform.\n\nFor lenders and partners, Autopilot MCP facilitates the development and deployment of AI agents customized to their specific workflows, guidelines, and borrower experiences, eliminating the need to rebuild underlying infrastructure.\n\nPrior to MCP, integrating AI into lending required separate connections for each system an agent needed to interact with. Mortgage processes, for instance, involve dozens of systems, including credit bureaus, pricing engines, underwriting platforms, title companies, compliance tools, and disclosure systems. These systems, often built over decades, were not designed for integrated operation, making each new connection a separate engineering, security, and compliance project.\n\nWith Autopilot MCP, any AI agent—whether developed by Blend, a lender, or a partner—can access Blend’s complete origination stack through a single interface, covering processes from credit and underwriting through compliance, disclosures, and closing. The platform introduces several core capabilities.\n\nThese capabilities include agentic workflow execution, allowing agents to perform lending workflows such as pulling credit, checking pricing, and verifying compliance, thereby preparing a complete, sequenced submission for loan officer decisioning. Institution-specific configuration enables each agent to operate against a lender’s own data, guidelines, and loan workflows within Blend, supporting portfolio products, HELOCs, or proprietary overlays alongside standard GSE guidelines.\n\nContinuous platform updates ensure that new Blend capabilities are automatically available to every lender with Autopilot activated, eliminating upgrade cycles or implementation projects. Additionally, built-in access controls log every agent action with a full audit trail, isolate credentials per deployment, and shut down access if a control layer is unreachable. Destructive operations, such as rate locks, credit pulls, and disclosure delivery, are gated until a lender is prepared to enable them.\n\nNima Ghamsari, Co-founder and Head of Blend, stated, “Until now, the hardest problem in lending AI wasn’t the intelligence of the models. It was getting them connected to the right systems, with the right controls, in a way a bank’s compliance team could actually approve. Autopilot MCP solves that. The intelligence is customizable, the infrastructure is shared. Lenders aren’t buying a specific AI feature. They’re getting a surface they can program.”\n\nLenders building on Autopilot MCP are positioned to extend and complement these capabilities as Blend Autopilot expands. Autopilot MCP is built on Model Context Protocol, an open standard for AI agent connectivity across the enterprise software industry introduced by Anthropic in 2024. This release marks the introduction of that standard to the lending sector.\n\nBlend Labs, Inc. (NYSE: BLND) operates as a digital banking solutions origination platform. Financial providers, including large banks, fintechs, credit unions, community banks, and independent mortgage banks, utilize Blend’s platform to enhance customer banking experiences.