MUFG backs Footprint's $25M plan to put AI agents on compliance cases

Why it matters

Bank compliance is becoming an early proving ground for vertical AI agents. MUFG's investment gives Footprint a path into institutions where governance and audit trails matter as much as model performance.

Footprint co-founders Eli Wachs and Alex Grinman announced on September 16th that Footprint raised a $25M Series B led by QED Investors, returning two years after QED led its Series A. MUFG, Commerce Ventures, LightBank and Alumni Ventures joined the round, alongside existing investors Index Ventures, Lerer Hippeau, BoxGroup, Operator Partners and Animal Capital.

Footprint will use the capital to expand Percy, its system of AI agents for financial-crime investigations, and Trust Fabric, the data and governance layer underneath it. Footprint also plans to double its engineering and sales organizations and open a San Francisco office, although it has not provided its current headcount.

The round advances a larger bet by Wachs and Grinman: identity verification can become the foundation for software that conducts the compliance work surrounding an identity. Footprint started in 2022 as a portable identity and secure data-vault product. Four years later, the founders are selling banks and fintechs a system designed to gather evidence, follow internal procedures, explain its reasoning and draft the record regulators may eventually inspect.

From portable identity to risk operations

Wachs began thinking about Footprint while studying economics and history at Stanford. He later worked on identity, privacy and security investments at General Atlantic. In a 2023 Fast Company profile, Wachs described an early version of Footprint as an "Apple Pay of identity," where consumers would verify themselves once rather than repeatedly hand sensitive data to separate businesses.

Grinman brought the security engineering. He studied computer science and cryptography at MIT and previously co-founded KryptCo, whose phone-based authentication technology was acquired by Akamai. On his personal site, Grinman says he worked across cryptographic protocol design, product, user experience and iOS engineering before starting Footprint.

Their original thesis addressed the tension between fraud prevention and onboarding friction. Footprint combined know-your-customer checks, fraud detection, authentication and encrypted storage, then tried to make a verified identity reusable. When Footprint raised a $13M Series A in May 2024, Wachs was still describing a network of portable identities and an onboarding suite.

The current product moves deeper into the work performed after a customer or transaction triggers a risk alert. Footprint introduced Percy as a beta AI assistant in March 2025, initially giving reviewers a way to question fraud signals, inspect triggered rules and act on user classifications. By December 2025, Grinman was describing Footprint as an AI-native system of record for onboarding, compliance and sensitive data.

That progression matters commercially. Identity verification supplies evidence. Compliance investigations determine how an institution interprets the evidence, applies policy and documents a decision. The latter puts Footprint inside a larger and more persistent operating budget, while exposing the product to much higher standards for consistency and auditability.

Percy has to show its work

Footprint says Percy can conduct anti-money-laundering reviews, enhanced due diligence, sanctions checks, know-your-customer and know-your-business workflows, and transaction-monitoring investigations. Percy accepts plain-English instructions, connects to sources including LexisNexis, Experian, ComplyAdvantage and government registries, and records citations, timestamps and each step taken during an investigation.

Trust Fabric is intended to retain the precedents produced by those investigations. A finding from one sanctions or due-diligence case can be applied to later work, while human reviewers approve proposed additions to that organizational memory. Footprint says agent changes are versioned and tested before deployment.

That architecture addresses the central problem with selling AI agents into regulated operations. Speed alone has limited value if a bank cannot reconstruct how an agent reached a conclusion, which data it used and whether it followed the bank's written procedures. Footprint is therefore pitching Percy as an auditable operating layer rather than a chatbot attached to an existing case-management system.

Footprint claims Percy has reduced some watchlist reviews from about 30 minutes to less than one minute. Footprint also says enhanced due-diligence work that previously took three hours can be completed in under 15 minutes while gathering 35% more evidence. Those figures are self-reported, and Footprint has not published enough deployment data to establish whether the results carry across institutions and case types.

Footprint identifies Bilt, Nuvei and MoonPay among its fintech customers and says its software is used by banks regulated by the Federal Deposit Insurance Corporation and Office of the Comptroller of the Currency. Footprint has not stated how many customers use Percy in production or how much of its business has moved from onboarding products to agent-led compliance work.

MUFG brings more than another venture logo

MUFG's participation is the round's clearest institutional signal. On September 17th, MUFG Innovation Partners said it invested through its third investment partnership after evaluating Footprint's data infrastructure and AI technology for financial-crime compliance.

MUFG Innovation Partners also said it would consider potential uses of Footprint across the MUFG group and other financial institutions. That language stops short of announcing a commercial deployment. It gives Footprint a route to test Percy against the requirements of a global banking group, where procurement, data controls and model governance can determine whether a promising agent reaches production.

Incumbent providers are pursuing the same opening. In May, FIS and Anthropic announced a financial-crimes AI agent designed to assemble evidence, evaluate anti-money-laundering cases and preserve a traceable record, with BMO and Amalgamated Bank among its initial institutions. Footprint must compete with vendors that already control core banking data and longstanding customer relationships.

Wachs and Grinman are betting that Footprint's narrower focus will let it build the risk workflow faster than large financial-software providers can modernize their installed systems. Footprint says Percy can work inside an institution's existing stack, reducing the need for a wholesale replacement. The technical work is only one part of that sale. Banks will decide whether Footprint's agents remain reliable as policies, sanctions lists, fraud patterns and regulatory expectations change.

The Series B brings Footprint's disclosed financing to at least $44M, including the $6M seed reported in 2023 and the $13M Series A. Footprint has not disclosed its valuation, revenue, pricing or retention.

QED's return and MUFG's strategic investment give Wachs and Grinman capital and institutional access for the next stage. Percy now has to prove that an AI agent can inherit a bank's procedures without inheriting the opacity that makes automated decisions difficult to defend.