Policy priorities for inclusive AI
The findings point to four practical priorities for firms, alongisde a need for clearer regulatory guidance.
As AI becomes more embedded in payment journeys, firms need to understand whether these services are working effectively for customers with characteristics of vulnerability. The research points to four practical priorities: better identification and monitoring, provision that responds to changing circumstances, straightforward access to human support, and inclusive digital identity.
1. Improve identification and outcomes monitoring
Firms need to understand where customers with vulnerability characteristics experience different outcomes. While 65% of respondents routinely monitor these customers separately, a significant minority do not.
Better identification and monitoring would help firms understand where problems arise, assess whether particular groups experience poorer outcomes and demonstrate how those issues are being addressed.
2. Design for changing circumstances
Provision is less consistent for customers experiencing recent adverse life events than for some other forms of vulnerability.
Firms should therefore consider how customer needs can change over time, rather than relying only on characteristics identified at the start of a relationship. Services should be able to respond when vulnerability emerges, or circumstances change.
3. Make human support easy to access
Most firms report that human support is available when an automated journey does not work as intended. Availability alone, however, is not enough.
Customers should be able to identify and access human support easily at the point of need. Firms should also monitor whether escalation routes work effectively and deliver appropriate outcomes.
4. Build inclusion into digital identity
Firms report high levels of technical readiness for digital identity, but inclusive provision is less developed.
As digital identity becomes more widely used, firms should consider vulnerable customers' needs from the outset. Verification processes should provide appropriate alternatives where customers cannot use a particular credential or digital route.
"AI can reach customers a branch network never could, but only if firms can prove it works for the people who need it most. The sector is not asking to be left alone; it is asking for a standard to build to."
Renuka Rawlins, Director of Policy and Government Relations, The Payments Association
Clearer regulatory guidance
Regulatory clarity and guidance is the most commonly cited barrier to inclusive AI adoption, identified by 47% of all 110 firms surveyed.
The findings point to demand for practical guidance on how existing regulatory expectations apply to AI-enabled customer journeys. Examples covering outcomes monitoring, inclusive design and access to human support could help firms translate existing requirements into practice.
Looking ahead
AI has the potential to improve access, personalisation and support across payments. The research also shows where further work is needed. Firms report high levels of confidence in their readiness, but reported practice and monitoring do not always support that confidence.
The priority is therefore to turn readiness into demonstrable customer outcomes. Better monitoring, provision that responds to changing circumstances, accessible human support and inclusive digital identity can help firms ensure that innovation works for a wider range of customers.