Dext Launches AI Connector to Bridge Financial Data with ChatGPT, Claude and Copilot

Dext has launched AI Connector, a Model Context Protocol (MCP) that integrates practice data with major LLMs like Claude, ChatGPT, and Copilot. For fintech professionals, this represents a significant shift toward "grounded" AI, allowing accountants to query real financial data using natural language without the risk of the generic hallucinations common in general-purpose chatbots.

What was announced

The AI Connector is designed to provide accounting and bookkeeping practices with a direct link between their Dext data and the AI tools they use daily. By grounding AI responses in a practice’s specific financial context, the tool aims to eliminate the guesswork often associated with general-purpose AI. Users can ask plain-language questions—such as identifying clients with the lowest health scores, checking live cash positions before meetings, or finding clients missing dividend accounts—and receive answers based on their actual Dext records.

The feature is built on Dext’s proprietary AI foundation, which has been trained over 16 years using more than 2 billion receipts and invoices. This accounting-specific training distinguishes it from general text models. Crucially, the AI Connector functions as a retrieval tool rather than a decision-maker; it is read-only and permission-aware. This ensures that users can only access data they are already authorized to see within the Dext platform, and the AI cannot make changes to the underlying records. The tool is currently available to Dext Advanced partners at no additional cost in the United Kingdom, Canada, the United States, Australia, and New Zealand.

"General-purpose AI is already part of how many practices work, but it doesn't know a practice's real numbers. AI Connector closes that gap: it puts the same Dext data our users already trust into the AI tools they're already using, so the answer they get back is grounded in their own practice, not a guess. This is an early step, and one we expect to build on."

Sabby Gill, CEO of Dext

The companies involved

Dext is a prominent provider of AI-powered bookkeeping automation, positioning itself as a critical layer in the digital accounting ecosystem. The company focuses on streamlining the "capture-to-pay" workflow, helping practices manage the vast quantities of financial documentation generated by their clients. With a history spanning over 16 years, Dext has built a significant data moat, having processed over 2 billion individual receipts and invoices to train its specialized accounting models.

The firm operates globally, with a strong presence in major English-speaking markets including the UK, North America, and Australasia. Its leadership team includes Chief Executive Officer Sabby Gill and Chief Product and Technology Officer Stephen Edginton. Dext’s strategy centers on supporting professional judgment through automation rather than replacing the accountant’s role. By providing tools like Practice Insights and team performance dashboards, the company has moved beyond simple data extraction into practice management and business intelligence, aiming to provide a comprehensive view of a firm’s operational health and client compliance.

What FF News has reported before

FF News has tracked Dext’s expansion into integrated financial services, notably with the report Dext Payments Launches in Canada: Streamlining Capture-to-Pay for QuickBooks and Xero Users. This followed a significant milestone where Dext Payments Expands to QuickBooks Online Following £6m Processing Milestone, highlighting the company's growth in the payments space.

Beyond product launches, FF News has covered Dext’s role in regulatory shifts. The company recently Appoints Former HMRC Lead to Navigate Making Tax Digital Transformation to help users manage the UK's tax digitisation. This move was timely, as Dext Data Reveals 65% of Accountants Missed First MTD for Income Tax Deadline, underscoring the ongoing need for automated compliance tools in the sector.

What this means

This announcement signals a move away from the "black box" era of AI in fintech. By utilizing the Model Context Protocol, Dext is challenging the industry to move beyond isolated chatbots toward integrated systems where AI acts as a sophisticated interface for existing databases. This puts pressure on traditional ERP and accounting software providers who have yet to open their data silos to third-party LLMs. The success of this approach will depend on how well practices manage the tension between AI-driven efficiency and the absolute necessity of data accuracy. It raises a critical question for the sector: will "grounded" AI finally make natural language the primary way accountants interact with complex financial data?