HELSINKI/GENEVA – May 27, 2026 – FinanceKey today announced the launch of a native Model Context Protocol (MCP) server, allowing enterprise AI tools such as ChatGPT, Claude and Gemini to securely access live treasury data directly. With the launch, FinanceKey becomes one of the first treasury platforms to bring MCP-native workflows to enterprise finance teams. 

MCP is the technology that allows AI tools to connect directly to external systems and read live data. While many treasury platforms are building closed AI experiences, FinanceKey has opened its platform to work with any AI tool an organisation chooses to use. This gives enterprise finance teams the flexibility to connect their existing AI tools directly to FinanceKey’s real-time data layer – while also laying the foundation for AI capabilities built into the platform itself. 

Built for real-time AI access 

Most treasury systems were not built for the AI era. They rely on overnight batch processing, scheduled exports and manual workflows, meaning AI tools often receive data that is already outdated by the time it is queried. 

FinanceKey was built differently. Designed API-first from day one, FinanceKey treats treasury data as a continuously updating operational layer rather than a series of static snapshots. Payments, balances, FX positions and forecasts update in real time as activity occurs across banks, ERPs and financial systems. 

The MCP server opens that live treasury layer directly to enterprise AI. 

That means a CFO can ask an AI assistant for a global cash position net of intraday movements and receive an answer in seconds – using current treasury data rather than yesterday’s exports. Treasury teams can surface liquidity exposures, analyse payment activity, draft reports or investigate exceptions directly through the AI environment their business already uses. 

“Most treasury platforms were built for a world of batch processing and end-of-day file-based statements. By the time the data reaches the people who need it, it’s already history,” said Veikko Koski, Co-Founder of FinanceKey. 

“AI only becomes operationally useful when it can access data that is accurate. Real-time infrastructure, and our API-first approach changes what AI can do inside treasury.” 

Open infrastructure, not a closed assistant 

Unlike AI implementations embedded inside individual software platforms, FinanceKey’s MCP server works as an open connection layer. Any MCP-compatible AI model or enterprise agent can securely query FinanceKey using the permissions and access structures already configured within the platform. 

“The industry is moving quickly toward AI agents and AI-native workflows, but most enterprise systems still aren’t accessible to them in any meaningful way,” said Macer Skeels, CTO and Co-Founder of FinanceKey. 

“MCP changes that. Instead of forcing customers into a single embedded assistant, FinanceKey allows organisations to connect their own AI ecosystem directly to live treasury operations. That creates a fundamentally more flexible foundation for how treasury teams will work in the future.” 

Security and governance 

Every AI query is subject to the same entity structures, user permissions and access controls already enforced within FinanceKey. If a user cannot access certain data inside the platform itself, they cannot access it through AI. 

The MCP server currently operates as a read-only connection. AI can retrieve information, perform analysis and draft outputs, but no actions are executed inside FinanceKey without human approval. Over time, FinanceKey plans to expand toward AI-assisted treasury workflows with human sign-off remaining mandatory at every stage. 

Available now 

For existing FinanceKey customers, no new implementation project is required. They can simply enable FinanceKey access in their enterprise AI tools of choice. The MCP server connects directly to the FinanceKey environment already in place, using existing APIs, data structures and permission models. 

Notes to editors

About FinanceKey 
FinanceKey is a real-time treasury orchestration platform that unifies cash visibility, payment operations, and forecasting across the entire finance organisation. Companies use  FinanceKey as an end-to-end payment and cash management solution or through its treasury API, feeding structured real-time data to data lakes, ERP, TMS or other systems. Trusted by global organisations managing billions, FinanceKey accelerates the shift to intelligent, connected treasury operations.

About MCP 
The Model Context Protocol (MCP) is an open standard developed by Anthropic that defines how AI models connect to external systems and data sources. It has since been adopted across the AI industry, enabling any compliant model to connect to MCP-enabled platforms without custom integration work. 

Media contact: Alex Cannon, Head of Growth +34 603 566 965
alex.cannon@financekey.com