Spendesk has launched an MCP solution enabling finance teams to query spend data via AI assistants, announced at Money 20/20 Europe.
The Spendesk MCP creates a conversational layer between the Spendesk API and AI assistants, including Claude and Dust, as well as other large language model interfaces. Finance professionals, including CFOs, financial controllers, and analysts, can submit natural language queries such as questions about current cash positions, pending invoices, or comparative spend across quarters, and receive responses drawn from live Spendesk data.
From static reports to real-time data access
According to the official press release, the underlying rationale for the product stems from a broader shift in how finance teams interact with data. Rather than relying on manually generated, static reports, companies are increasingly using AI tools to accelerate decision-making. Spendesk's own platform data reflects this: AI tool spend among its customers has grown more than 50-fold since 2022, with year-on-year growth of 281% recorded in early 2026. AI tools now account for 38% of all technology spend processed through the platform, compared with under 1% four years ago.
The MCP is positioned as a retrieval and analysis layer, meaning it is designed to surface data and insights without granting AI assistants control over sensitive workflows such as payment approvals. This separation is intended to preserve governance and oversight over financial operations while extending access to spend intelligence across teams.
Quentin Vigneau, CPO of Spendesk, described the strategic intent as positioning Spendesk as a foundational platform through which agents can operate in an agentic era, while keeping engineering focused on core user experiences. The MCP is framed as part of a broader set of agentic initiatives rather than a standalone feature.
Availability and access
The solution is currently available in beta. Customers seeking early access can request activation through Spendesk directly.
The launch reflects a growing pattern among fintech platforms of adopting MCP as a standard integration layer, allowing proprietary data to be made accessible to AI tools without requiring custom-built connectors for each assistant. For spend management platforms, this has particular relevance as finance teams consolidate around AI-native workflows and demand interoperability between financial data systems and the tools they already use.
