The Rise of B2AI: How Visa and InFlow are Enabling AI Agentic Payments
The era of AI agents performing complex tasks is already here. From researching humanoid robot ecosystems to managing calendars and emails, agents are increasingly capable of executing high-level workflows. However, a critical friction point has remained: the inability of these agents to independently activate or pay for the services they require. That barrier is now coming down.
San Francisco-based InFlow has launched an agent-native commerce infrastructure built on Visa Intelligent Commerce. This partnership introduces a structural shift in how transactions occur, moving beyond human-to-machine interactions toward a new economy: Business-to-AI (B2AI).
- B2AI Infrastructure: InFlow is building a system where the primary customer is an AI agent rather than a human.
- The Policy Engine: The system decouples the decision to buy from the act of payment, allowing humans to set strict rules for AI spending.
- Visa’s Role: Visa Intelligent Commerce provides the essential trust layer, including tokenization, authentication, and global merchant acceptance.
- Market Demand: According to Visa’s Business-to-AI report, 71% of businesses are willing to optimize products specifically for AI agents.
Solving the “Intelligence vs. Access” Gap
For a long time, the bottleneck in agentic commerce wasn’t a lack of AI intelligence, but a lack of AI-native payment systems. AI agents could identify a need or a service, but they couldn’t execute the financial transaction to secure it.
Jim Nguyen, founder of InFlow and a former PayPal executive, notes that while agents are getting smarter, they still struggle to activate services on their own. InFlow positions itself as the “PayPal for agents,” creating a bridge that allows AI to operate within a secure financial framework.
Decoupling Decision from Entry: The Policy Layer
The most significant innovation in this launch is the introduction of a policy-governed payments engine. In traditional commerce, the “decision layer” (deciding what to buy and how much to spend) and the “entry layer” (inputting card details) are fused into a single human action. When a human is removed from the loop, that policy usually disappears with them.
InFlow addresses this by providing the AI agent with a wallet that comes with built-in rules. This prevents the risk of a “dumb” wallet being drained by any entity with the right credentials. Instead, the policy engine dictates:
- What the agent is allowed to spend on.
- Who the agent can pay.
- What limits are imposed on specific transactions.
The Trust Layer: Why Visa is Critical
While some might look toward cryptocurrency or stablecoins for autonomous payments, the partnership with Visa anchors agentic commerce in the existing global financial system.
Visa Intelligent Commerce provides the necessary infrastructure to make these transactions viable in the real world. By supplying payment credentials, tokenization, and authentication, Visa ensures that an agent transacting through InFlow is using credentials that any merchant on Visa’s global network can recognize, and accept.
“AI agents are a new buyer category, and businesses need a trusted way to support them,” said Visa VP Tanner Riche. “Visa Intelligent Commerce helps enable secure, trusted credentials for agent-initiated transactions.”
Targeting the B2B Cloud Ecosystem
While the long-term vision includes consumer-facing agentic commerce—such as an agent purchasing a specific pair of shoes when they hit a certain price point—the immediate focus is B2B cloud infrastructure. This includes the essential layers agents consume to function, such as:

- Compute and inference power.
- Data and storage services.
- Specialized capability layers.
This focus aligns with broader corporate trends. Visa’s Business-to-AI report indicates that 77% of businesses are already using or piloting AI in their operations, signaling a massive emerging market of companies ready to sell to non-human customers.
The Future of Autonomous Commerce
The launch of InFlow’s infrastructure marks a transition from AI as a tool for productivity to AI as an economic actor. By solving the structural problem of payment policy and leveraging the global reach of the Visa network, the industry is moving toward a world where B2AI is a standard category of commerce.
The ultimate success of this model will depend on how quickly other digital payment players adapt and whether businesses can successfully shift their sales and optimization strategies to cater to a buyer that doesn’t have a human pulse, but does have a budget.