AI in Payments: Efficiency vs. Strategic Orchestration

Finextra and Volante hosted a recent webinar where industry experts discussed how firms are using artificial intelligence in payments and the governance questions that follow.

Panelists noted AI already delivers measurable benefits for banks and processors, helping reduce manual work, improve exception handling and strengthen fraud detection.

Speakers warned that limiting AI to back-office tasks could leave broader opportunities unused as payments increasingly move across multiple networks, rails, currencies and tokenised assets. They described a future in which any single transaction can travel several possible routes through traditional banks, real-time schemes and newer channels such as stablecoins and central bank digital currencies.

The panel outlined a use case they called dynamic payment orchestration, where AI evaluates factors including transaction value, currency, corridor, speed, cost, liquidity and risk to select the most suitable route for each payment in real time. Panelists argued manual decision making will not scale in such an environment.

Discussion also focused on accountability and regulatory risk. Panelists raised the question of who should be responsible when an AI-recommended route results in financial loss, excessive delay or a compliance breach. They highlighted gaps in explainability and oversight when models move beyond support roles into decision-making functions.

To address those gaps, the panel proposed several operational changes. Suggestions included keeping humans in the loop for high-risk decisions, creating audit trails that record AI recommendations and actions, and building monitoring systems to detect model drift or unintended outcomes. Panelists said regulators and compliance teams will expect such safeguards as AI becomes more embedded in routing and settlement decisions.

Nadish Lad, global head of product and strategic business at Volante Technologies, participated on the panel. Teresa Connors, a contributing editor, moderated the discussion. They emphasised that firms must combine technical change with stronger controls and clearer regulatory engagement to maintain transparency and meet supervisory expectations.

Panelists concluded the industry faces a near-term choice: continue applying AI mainly to reduce costs and speed operations, or invest in systems, governance and skills that extend AI into payment routing and settlement. They said that decision will affect whether AI only accelerates existing processes or also changes how value moves across payment networks.