Net 30 has endured through a decade of invoice digitisation, bank connectivity and more sophisticated cash forecasting, but the arrival of agentic artificial intelligence is beginning to expose how little has changed in the economics of business-to-business payments.
The core promise is not simply faster accounts payable. It is a shift in which payment terms stop functioning as fixed clauses written months earlier and start behaving more like live prices for working capital. In...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
that world, software could assess whether a buyer should release cash earlier than agreed, while a supplier’s system works out the discount that makes immediate payment worthwhile.
That future remains largely theoretical. JPMorgan outlined a similar model in June, imagining treasury agents on both sides of a transaction negotiating timing and discounts within preset limits. But the bank’s vision depended on several prerequisites that do not yet exist in most firms: standardised interfaces, interoperable policy frameworks and enough trust between counterparties for machines to bargain on their behalf.
Even so, the direction of travel is becoming clearer. Mastercard has argued that agentic AI in commercial payments will depend on trust, control and redesigned workflows, with governance, accountability, liability and auditability at the centre of adoption. The International Monetary Fund has made a similar point, warning that autonomous systems must operate inside payment infrastructures built for deterministic outcomes, not probabilistic ones.
That tension is exactly why the practical challenge is less about the chatbot interface and more about the plumbing underneath it. A machine deciding whether to pay an invoice 11 or 14 days early would need reliable visibility into cash balances, forecast obligations, supplier identity, contractual restrictions and the return the company could earn by keeping the money longer. In other words, the system would need certainty about reality before it could even begin to make a rational offer.
Industry surveys suggest finance teams are already moving towards this more data-driven operating model. Visa, in a report produced with PYMNTS Intelligence, found that seven in 10 adaptive chief financial officers and treasurers are using working capital tools to pay suppliers more quickly, preserve flexibility and strengthen commercial relationships in volatile conditions. PYMNTS Intelligence also reported that 58 per cent of small and medium-sized businesses consider integration very or extremely important when assessing new technology, underscoring how central connected systems have become to any serious automation strategy.
The policy question may prove even more important than the technology question. Treasury rules were written for people, not autonomous agents, and they will need to be translated into machine-readable boundaries if AI is to do more than suggest actions. A firm might allow an agent to accelerate payment only when the discount clears a specified hurdle rate, or forbid it from changing terms for strategically important suppliers. It might also require human sign-off whenever a proposed move materially alters monthly liquidity.
This is why many observers expect adoption to arrive in stages rather than through a sudden handover of authority. The first phase is likely to be AI identifying opportunities in working capital. The next is execution of tightly constrained decisions. Only after that, if the surrounding controls mature, would agents start negotiating directly with each other across companies.
PYMNTS has argued that the most radical element of agentic accounts payable is not the removal of humans from invoice processing, but the creation of liquidity inside every invoice. Mastercard’s analysis points to the same conclusion from a different angle: if AI is to participate in commercial payments, it must do so inside deterministic guardrails and with a hybrid model that preserves human oversight.
The broader implication is that business payments may be moving away from static terms and towards continuously priced decisions. Instead of treating payment timing as an administrative afterthought, finance teams may increasingly see it as a live portfolio choice, one in which every payable represents a potential financing trade-off.
For now, net 30 remains the default. But if agentic AI can reliably compare liquidity, risk, return and counterparty need in real time, the old language of fixed terms may begin to look less like a rule of commerce and more like a temporary compromise.
Source: Noah Wire Services
