Finextra AI payments webinar 2026 to examine automation and accountability

Finextra’s AI payments webinar 2026, held in association with Volante Technologies, is scheduled for 21 October and will examine the move from operational automation to AI-assisted payment decisions. The online session will consider potential efficiency gains in fraud detection and exception handling alongside the governance questions raised when AI influences payment routes or liquidity management.

The event begins at 15:00 BST, 16:00 CEST and 10:00 EDT. Nadish Lad, global head of product and strategic business at Volante, is listed as a speaker, while Finextra contributing editor Teresa Connors will moderate.

Finextra says financial institutions are using AI to streamline operations, reduce manual intervention, improve the handling of exceptions and strengthen fraud detection. The organiser presents these as current applications, while positioning broader changes in payment decision-making as a longer-term prospect.

What could AI do in payment routing and liquidity management?

Finextra describes a possible future in which organisations must choose among traditional banking infrastructure, real-time payment schemes and other routes. It says an AI system could assess a transaction’s value, currency, corridor, speed, cost, liquidity and risk to identify a suitable route. The organiser calls this potential use dynamic payment orchestration.

That prospect concerns more than processing a payment faster. It would put AI into decisions that may determine how funds are routed or which payments receive priority when resources are constrained. Finextra frames responsibility for losses, delays and compliance issues arising from an AI-recommended route as a central question for the industry, alongside oversight, transparency and trust.

A November 2025 working paper by Bank for International Settlements and Bank of Canada researchers provides a limited test of that possibility. Using prompt-based experiments with ChatGPT’s o3 reasoning model, the authors simulated high-level intraday liquidity management in a stylised wholesale payment system, including liquidity shocks and competing payment priorities.

The study found that the tested agent closely replicated key prudential cash-management practices and generated recommendations intended to preserve liquidity while limiting delays, despite receiving no domain-specific training. The results did not involve a live deployment or a particular payment system, and the paper notes that real-time gross settlement arrangements vary across jurisdictions. The authors say routine cash-management work could potentially be automated, subject to regulatory and policy safeguards.

Intraday liquidity management requires firms to balance funds held in advance against settlement timeliness. More pre-funded liquidity can reduce delays but carries an opportunity cost, while less liquidity can lower that cost but can leave payments queued for longer, according to the BIS paper.

The governance discussion extends beyond payment routing. A December 2025 article in the Journal of Financial Stability identifies opacity in machine-learning models, reliance on extensive data, privacy, cybersecurity and algorithmic-bias risks. It also points to possible concentration, model-herding and network-connectedness concerns as AI use develops in finance.

The distinction for payment operators is therefore between routine automation and systems that advise on liquidity, prioritisation or routes with potential settlement and compliance consequences. Finextra’s programme places that boundary, and the accountability around it, at the centre of its October discussion.

This story draws on original reporting from Finextra Research.