Summary

• Pliant has raised $80m in Series C funding led by Seaya Nova and Speedinvest, with three additional investors participating.
• The company's AI integrations allow agents to access payment information and assist with administration through Model Context Protocol.
• Current integrations do not independently move money or issue cards; more extensive agentic payment functionality remains part of the development roadmap.

Berlin financial technology company Pliant has raised $80 million to expand its business payments infrastructure and develop financial services designed to work with artificial intelligence agents, as automated software begins handling more of the administration surrounding commercial transactions.

The Series C financing was led by Seaya Nova and Speedinvest, with participation from Illuminate Financial, Motive Partners, and Molten Ventures. The investment will support further expansion in the United States and Canada, entry into additional international markets, and development of the company’s AI capabilities.

Pliant operates a modular platform combining commercial card issuing, payment accounts, foreign exchange, and expenditure management. It provides these services directly to businesses and through banks, software companies, and financial technology providers building their own payment products.

The company says annual recurring revenue has exceeded $100 million after three consecutive years of doubling its business. It employs more than 300 people and supports commercial card programmes across numerous international markets, although detailed independently audited revenue figures were not included in the latest funding material.

Agents gain access to financial workflows

Business finance involves substantial administrative work surrounding payments, including checking transactions, collecting receipts, assigning expenditure to the correct accounting categories, and reconciling records with invoices.

These processes often require employees to move information between several applications, particularly where payment systems, accounting platforms, and purchasing software are not fully integrated.

Pliant is extending its existing application programming interfaces through Model Context Protocol, allowing external AI assistants to interact with financial information and perform supported administrative actions.

The company’s documentation identifies integrations with tools including Claude, ChatGPT, Cursor, and other compatible systems. Users can ask agents to retrieve transaction information, examine expenditure, update accounting classifications, and assist with receipt management without moving manually between applications.

Crucially, the currently documented integrations do not authorise AI agents to move money autonomously. Pliant states that money movement and card issuance are not exposed to its agent tools today, while actions that change financial information require confirmation from an authorised user.

This distinction is important because retrieving information and executing a payment create different operational responsibilities. An assistant that identifies a missing receipt can help resolve an administrative problem without being given authority to transfer funds or change the company’s financial exposure.

As businesses consider delegating more work to AI systems, they must determine which actions can be performed automatically and which require a person to approve the outcome.

Pliant’s longer-term agentic payments offering is intended to extend the range of supported activities while preserving budgets, approval limits, and individual accountability. The company has not established that unrestricted autonomous payment execution is generally available.

Existing permissions remain central

Pliant’s agent integrations operate under the permissions associated with individual users. A connected assistant should therefore receive access only to information and functions that the responsible employee is authorised to use.

The arrangement also provides a route for recording activity and establishing who requested a particular action. This can be particularly useful where AI assistants operate across different business applications and perform several administrative steps during a single task.

For example, an agent might retrieve an invoice from email, compare its details with an existing card transaction, and prepare an accounting update. Each operation relies on information from a different system, while the final change needs to remain consistent with the company’s financial controls.

However, an application following a user’s permissions does not automatically guarantee that every proposed action is appropriate. An employee may possess authority to change a transaction classification but still need to verify that the agent has selected the correct category.

The same problem becomes more significant when financial automation extends towards payments. A system could correctly identify an invoice while making an incorrect judgement about whether it has already been paid, whether the supplier’s details are authentic, or whether the transaction complies with purchasing rules.

Controls must therefore address the quality of underlying information as well as the authority to perform an action. Human confirmation reduces some risks but remains effective only when the approving employee has sufficient information to assess the proposed transaction.

Regulation shapes the market opportunity

Pliant’s underlying payment business operates through regulated financial infrastructure, giving it a different role from software companies that provide AI assistants without directly delivering financial services.

The group holds a European electronic money licence through its Finnish entity and has established partnerships supporting operations in other markets. It is also a principal member of Visa and Mastercard, placing it within the regulated infrastructure used to issue and manage commercial payment instruments.

In October 2026, Pliant obtained UK electronic money regulatory approval intended to support the provision of services directly through its regional subsidiary. The company also entered the United States through a partnership with Coastal Financial.

These arrangements help Pliant offer products in markets with different legal and regulatory requirements, although licensing in one jurisdiction does not automatically provide permission to operate every financial service elsewhere.

As AI systems become more closely integrated with business finance, the organisations providing regulated payment infrastructure will still need to maintain safeguards covering financial crime, fraud, customer protection, operational resilience, and transaction records.

Vinoth Jayakumar, general partner at Seaya Nova, identified that responsibility in explaining the investment: “When agents move money, someone still has to be licensed, auditable and liable, and that is the layer Pliant owns.”

The observation describes the commercial rationale for developing agent-compatible payment infrastructure, rather than confirming that Pliant’s existing AI integrations already execute unrestricted transfers.

Growth depends on practical integration

Pliant’s business spans direct corporate customers and organisations embedding payment services into their own software. Its customer base includes travel, insurance, and technology companies, where payments can involve numerous suppliers, currencies, and reconciliation processes.

These industries provide potential applications for AI-assisted administration because employees frequently work with structured financial information distributed across several systems.

Automation could reduce repetitive processing where records are accurate and workflows are clearly defined. The benefits become less certain when invoices contain incomplete information, supplier arrangements differ between transactions, or exceptions require commercial judgement.

The company’s modular platform may allow customers to introduce selected AI capabilities without replacing their existing payment arrangements. However, organisations will still need to assess integration costs, data handling, permission management, and the reliability of automated outputs.

For banks and software providers offering Pliant-powered services under their own brands, agent compatibility also introduces questions about responsibility when a customer uses a third-party assistant to interact with a financial product.

The new financing gives Pliant additional resources to develop those capabilities while expanding geographically. Its existing revenue and international payment infrastructure provide a commercial base, but the eventual demand for agentic payments will depend on how confidently organisations can delegate financial work without weakening established controls.

The immediate product capability is better described as AI-assisted financial administration with human approval for consequential actions. More extensive automated payment execution remains an area of development, with commercial adoption dependent on the security, auditability, and regulatory treatment of the resulting workflows.