Agent-to-agent B2B billing forms a category with a $20M seed
Agents that negotiate invoices, settle payments and resolve billing disputes with the counterparty's agents. Finance teams pay.
Competitive · Causa Prima $20M seed
Written for: Investor or analyst
Next step
Thesis: B2B billing moves to agent-to-agent negotiation, and the network that sets the protocol wins. Confirming signal: an ERP or AP platform announces agent-to-agent interoperability, or Causa Prima names large counterparties on both sides. Already building: Causa Prima, Stripe (AI-readable pricing), x402 middleware projects.
Why now
A seed round this size for agent-to-agent finance shows investors are pricing a category, not a feature. Billing disputes are slow, manual and costly in accounts receivable, which gives a clear ROI. It is still too early: two sides have to adopt for it to work, so network effects decide the winner and there is no standard protocol yet.
What became possible
Causa Prima raised a $20M seed for a platform where one company's agents deal directly with another's on B2B payments and billing disputes. Stripe's Monetize 2026 sessions describe pricing that has to be readable by AI buyers, and x402 payment middleware is spreading.
One way in
- Dispute resolution network
A neutral protocol and ledger where buyer and supplier agents trade evidence and settle invoice disputes under pre-agreed rules.
- Customer
- AR teams at mid-market suppliers with high dispute rates, such as logistics and distribution
- Model
- Per-resolved-dispute fee charged to both sides
Competition
The hard part is cold start. A neutral network needs both sides on board, and Causa Prima and Billtrust already have two-sided networks. So the entrant has to win a vertical where one big shipper or retailer can require its suppliers to adopt the rules. Otherwise it risks becoming a feature that these networks absorb.
Full competition notes (10 competitors)
The space is taken at the top. Causa Prima announced its $20M seed on October 13, 2026, and already runs agents on both the buyer and supplier side that settle invoice disputes, with more than 3,000 accounts on its network. Incumbents (HighRadius, Genpact, Billtrust) and logistics specialists (Shipsy Vera, Lighthouz) already sell dispute and deduction agents to AR teams. Still, a new entrant has room in two places. First, neutrality. Causa Prima owns its network and makes money from early-payment cuts, so it is a party with its own economics, not a neutral referee. The neutral-protocol work so far (the AAA/Integra Legal Context Protocol, the Consulate ADP and ADRP IETF drafts, the W3C agentic arbitration group) is general-purpose, still in draft, and has no evidence formats specific to invoices, such as proof of delivery, rate confirmations, detention and accessorial charges, short-ship claims or retailer compliance chargebacks. Second, vertical depth. Logistics and distribution disputes depend on carrier documents, telematics and EDI 210/214 data that a horizontal finance network is unlikely to model well. An entrant could build the dispute rulebook and evidence ledger for freight and distribution, connect to agents from any vendor (Causa Prima, HighRadius, buyers' AP tools), and possibly ride on the LCP standard and AAA for escalation. The hard part is cold start. A neutral network needs both sides on board, and Causa Prima and Billtrust already have two-sided networks. So the entrant has to win a vertical where one big shipper or retailer can require its suppliers to adopt the rules. Otherwise it risks becoming a feature that these networks absorb.
- Causa Prima
This is the $20M seed in the opportunity title. Base10 led the round, announced October 13, 2026, and the company has raised $30M in total including a Creandum-led pre-seed in June 2026. It runs agents on both sides of a deal, the buyer's AP and the supplier's AR, which settle invoice disputes and negotiate payment terms with each other. It reports more than 3,000 active accounts. It earns a cut of the early payments its agents negotiate, so the dispute work mainly brings users into a payments-terms business. Sources disagree on where it is based (Madrid/Munich vs. Delaware), and it is opening a New York office.
- Shipsy Vera
An autonomous agent that resolves freight disputes with carriers and vendors. Shipsy says it has resolved over $25M in disputes for a global drinks company, named as Heineken. It sits on one side of the dispute only, not on a neutral network, but it already targets the logistics vertical.
- Lighthouz AI (YC S24)
Audits FTL/LTL freight bills, files disputes and automates AP/AR and collections for freight brokers. It works for one side only, in the logistics vertical.
- HighRadius Deductions
The incumbent in AR deductions and disputes. It sells agentic AI that predicts whether a deduction is valid, pulls documents and runs dispute workflows. It works only for the supplier and does not use agent-to-agent protocols.
- Genpact Agentic Deductions
An agentic, services-led deductions product that finds, checks and resolves disputed deductions for consumer goods suppliers.
- Billtrust Business Payments Network
Connects suppliers' AR systems directly to buyers' AP portals. This existing two-sided network could add agent-based dispute handling.
- Legal Context Protocol (AAA + Integra Ledger)
A draft open standard from June 2026, backed by the American Arbitration Association, with Google, IBM, Circle and Wayfair reported as contributors. It lets agents find a merchant's terms, record acceptance and see the dispute route in machine-readable form. It covers the neutral pre-agreed-rules layer but not the exchange of invoice evidence.
- Consulate / Agentic Dispute Protocol
An individual IETF draft (October 2025) plus a W3C Agentic Arbitration community group. It defines machine-readable messages for filing a dispute, submitting evidence and issuing an award, with Consulate, Inc. as the reference implementation. It is early, general-purpose and not built for invoices.
- ADRP IETF draft
Another individual IETF draft for agent-to-agent dispute resolution. It separates disputes that code can settle from a proof bundle from those judged against pre-agreed acceptance criteria, with escalation to arbitration.
- Pactum
Negotiation agents on the buyer side that trade payment terms and early-pay discounts with suppliers, including a Walmart deployment. They are next to dispute work and could grow into it.
Checked 2026-10-14
Risks
- Needs both buyer and seller on board, so the cold start is hard
- ERP vendors (SAP, Oracle) and AP platforms (Bill) can add agent endpoints themselves
- No liability framework yet for agent-agreed settlements
Best on team fit; buildability is the weak point (4).
tech 5 · pull 6 · space 6 · fit 7 · build 4
Exposed
Companies that could benefit or be threatened. News and context, not investment advice.
Momentum of its main topic
Sources
Status history
- 2026-10-14 Added to the watchlist