The Pulse
Mollie Brings 60+ Financial Workflows to Claude and ChatGPT
Mollie says businesses can prepare more than 60 financial operations through AI assistants, including payment links, subscription changes and draft invoices. Its plugins and open MCP server connect to Claude and ChatGPT, with human approval
AI.info Team · Written with AI · fact-checked ·
More than 60 financial operations are now available through AI assistants connected to Mollie, the European financial services provider announced on October 29. Its official plugins and open Model Context Protocol (MCP) server let businesses use Claude and ChatGPT to prepare tasks such as creating payment links, managing subscriptions and drafting invoices. Mollie says a person remains in control of each transaction.
From payment links to draft invoices
The integration reaches beyond questions about account activity. Mollie describes assistants preparing multi-step actions across its services, including checkout links, subscription changes and invoices, as well as reporting on payment methods and average order value. Users can phrase requests in ordinary language rather than finding the relevant page and entering details in Mollie’s dashboard.
For example, a business could ask for a payment link with a specified amount and expiry, or request a draft invoice with a due date. Mollie says its system handles the preparation, while the user makes the decision. The company does not detail in its announcement how the more than 60 operations are divided among read-only queries, draft creation and actions that change account data.
OAuth connects the assistant to Mollie
Mollie says businesses can connect through an authorization screen using OAuth, without code, API keys or manual API setup. The company puts setup at less than a minute and says users can review which capabilities the assistant is allowed to prepare. Each team member can connect their own assistant, according to Mollie’s product explanation.
The approach brings payment and account work into the same conversational tools some teams already use for other tasks. Mollie says the plugins are available through major AI marketplaces and developer environments, and that the MCP server is built on an open standard. The company presents that standard as a way for businesses to keep their choice of assistant separate from their choice of payment provider.
Human approval remains the boundary
Mollie says an assistant cannot move money by itself. For a transfer request, it can prepare the transaction and check recipient details, then stop for a user to review and approve the summary. The company says existing account roles, daily limits and security controls continue to apply.
“Entrepreneurs and finance teams want to spend their time growing their business, not on admin.”
— Rogier Schoute, Chief Product Officer at Mollie
Schoute says the assistant handles preparation while the business owner stays in control. That distinction matters when a chat prompt can trigger work in a payments account: the model can assemble an action, but approval remains a separate step.
The assistant prepares; the business decides
Mollie’s launch shifts its AI connection from a tool aimed chiefly at developers toward ready-to-use plugins for business and finance teams. Its product explanation describes the target users as founders, operations leads and finance staff who handle routine work across multiple screens. The company is now packaging those actions for people who may not work directly with APIs.
The release leaves practical questions for businesses to assess as they adopt the tools: which permissions to grant, which tasks to allow an assistant to prepare, and how carefully employees review proposed changes before approving them. Mollie’s stated safeguard is specific: money does not move until a person signs off.