Payment rails run on an assumption so old it rarely gets named: the entity authorizing a transaction is a human being. Account credentialing and fraud scoring both trace back to a cardholder or a registered business with a human accountable at the end of the chain. An AI agent acting autonomously has no assigned place in that structure. It can be instructed to transact, but the infrastructure underneath was not written for a non-human principal. San Francisco-based Natural is building that missing layer, and on July 20, 2026, the company announced a $30 million Series A led by Kirsten Green at Forerunner, with continued participation from all prior investors.
The mechanism behind the gap
Where this sits in the stack: card network rulebooks and ACH operating rules define authentication in terms of human account holders who can be credentialed and held liable. An agent operating inside those rails either borrows a human's credentials, which introduces security and compliance exposure, or runs into walls at the authorization step. The infrastructure question is not how to make agents faster at payments. It is how to create a principal type that existing settlement and routing systems will accept.
193 days old, $30 million raised
The round closed when Natural was 193 days old. That pace reflects where institutional attention has moved. Forerunner, led by Kirsten Green, anchored the Series A, and all major investors from prior rounds continued participation, though the announcement does not name them. Total capital raised is not disclosed beyond the $30 million from this round.
Natural describes itself as building payments infrastructure for AI agents, a layer that other builders integrate rather than a direct consumer product. The $30 million Series A is the first large institutional commitment the company has disclosed publicly.