The proposal comes at a time when net‑30 and net‑60 terms still dominate corporate cash‑flow arrangements, even though companies have largely digitized invoices, automated accounts‑payable workflows and linked bank accounts through APIs.
JPMorgan’s sketch envisions each party’s treasury agent accessing machine‑readable policies, clean data and hard authority limits, then using AI to calculate whether paying an invoice early—say 11 days ahead—yields a better return than holding the cash, while the supplier’s agent determines the discount that makes immediate payment worthwhile.
If both agents agree, the payment timing and discount would be negotiated in real time, converting a static contractual provision into a dynamic price that fluctuates with liquidity, interest rates, counterparty risk and the company’s short‑term investment opportunities.
The vision, however, remains prospective. No implementation exists yet, and it would require standardized interfaces, interoperable treasury policies and a high degree of trust between counterparties before autonomous agents could act without human sign‑off.
Industry data underscores why such a shift matters: a Visa‑PYMNTS Working Capital Index for 2025‑2026 found that seven in ten “adaptive” CFOs and treasurers already use working‑capital solutions to pay suppliers faster and improve agility in a volatile economy.
Experts caution that before AI can negotiate, finance teams must first automate certainty. An AI decision engine needs reliable cash forecasts, validated supplier identities, contractual restrictions and a clear picture of alternative returns on retained cash.
Corporate treasury policies, traditionally written for people, would have to be translated into executable rules. For example, an agent might be authorized to accelerate payments up to a set amount when the expected early‑payment discount exceeds a hurdle rate, but barred from extending terms for strategically critical suppliers.
The near‑term trajectory is expected to move from AI‑identified working‑capital opportunities, to recommendation engines, then to narrowly defined execution, and finally to full‑blown autonomous negotiation with external agents.
If realized, the shift could transform every payable into a real‑time decision about the highest‑value use of corporate cash, reshaping supplier relationships, liquidity management and the broader B2B payments landscape.