Notifications
Clear all
Topic starter
10/07/2026 9:12 pm
TL;DR: Agentic payments on Base crossed 100 million transactions in roughly three quarters, with $1+ transfers rising from 49% to 95% of volume and tester-to-payer conversion improving 4x, according to Chainalysis. The governance challenge is no longer whether autonomous systems will need payment rails, but how identity, authorization, and compliance controls adapt when machines transact at speed.
NHIMG editorial — based on content published by Chainalysis: The New Rails: How Digital Assets Are Reshaping the Foundations of Finance
By the numbers:
- Agentic payments on Base crossed 100 million transactions in approximately three quarters, surging from near-zero in Q3 2025 before moderating in early 2026.
- Transactions of $1+ now represent 95% of total volume transferred, up from 49% in early 2025.
Questions worth separating out
Q: How should security teams govern AI agents that can spend money autonomously?
A: Treat spend authority as a controlled entitlement, not a convenience feature.
Q: Why do machine-to-machine payment flows create new IAM challenges?
A: Because the control problem shifts from authenticating a person to governing a machine that can initiate value transfer, resume access, and act repeatedly at speed.
Q: What do organisations get wrong about agentic payment approval?
A: They often assume one approval is enough for an ongoing machine workflow.
Practitioner guidance
- Define agent payment authority as a distinct entitlement class Create a separate approval model for machine spend that records which agents can pay, what they can pay for, and the transaction ceilings that apply.
- Bind wallet activity to service ownership and business purpose Require each payment-capable wallet to map back to a named workload, owner, and use case so investigators can reconstruct why a transfer occurred and whether it was within policy.
- Use short-lived policy scopes for machine spending Limit autonomous payment permissions to time-bound, context-bound tasks instead of persistent wallet authority, and revoke access automatically when the task or model session ends.
What's in the full report
Chainalysis' full blog covers the operational detail this post intentionally leaves for the source:
- Quarter-by-quarter transaction breakdowns showing how Base activity moved from experimentation to sustained usage
- Wallet-age, balance, and token-diversity comparisons that help teams understand the user profile behind agentic payment adoption
- Tester-to-payer conversion and weekly retention data that show whether usage is becoming durable
- Protocol-level examples of how x402 handles payment authorization within the request flow
👉 Read Chainalysis' analysis of agentic payments on Base and x402 adoption →
x402 and agentic payments: what it means for IAM and governance?
Explore further
10/07/2026 9:35 pm
Agentic payments create a new identity class that IAM has to recognise. A machine that can request, pay for, and resume access is not just an application client. It is a governed actor with its own entitlement scope, audit trail, and failure modes. That shifts control design away from static credentials and toward lifecycle-managed machine identities with explicit payment authority. Practitioners should treat payment-enabled agents as a distinct policy domain, not an edge case.
A question worth separating out:
Q: Who is accountable when an autonomous agent makes an unauthorised payment?
A: Accountability should sit with the service owner and the business approver who granted the spend authority, supported by security and compliance controls that recorded the policy decision. If no one can show the approved purpose, scope, and transaction history, the governance model is incomplete.
👉 Read our full editorial: Agentic payments are pushing AI agents into financial operations
