Progress Together has submitted evidence to the Financial Conduct Authority’s Mills Review, urging regulators and firms to address the risk that artificial intelligence could unintentionally widen socio-economic inequality in retail financial services.
The submission, responding to Theme 3 of the consulation: Future Consumer Trends, highlights how AI is already shaping financial services – from credit decisions and fraud detection to customer support and digital onboarding.
While these technologies promise faster decisions and more personalised services, Progress Together’s evidence warns that without deliberate safeguards, AI systems could embed and scale existing inequalities experienced by lower socio-economic consumers.
The FCA’s Mills Review is examining how technological change may affect consumers in the coming decade. Progress Together’s response focuses on how data patterns, workforce diversity, and design choices could shape whether AI improves or restricts financial access.
When AI learns from unequal systems
AI systems learn from historical data. In financial services, those data patterns often reflect the behaviours and circumstances of consumers with stable incomes, strong digital access and conventional credit histories.
For many consumers from lower socio-economic backgrounds, financial lives look different: irregular income patterns, thinner credit files, limited digital access and more complex household financial pressures.
These differences can be misinterpreted by automated systems as signs of higher risk.
Without careful design and oversight, AI may therefore amplify existing barriers, limiting access to credit, financial products or support services for some consumers.
Workforce diversity also shapes outcomes
Progress Together’s evidence also highlights the workforce dimension of AI development.
Research from its Rise with AI report shows that access to AI tools and training is already uneven across the financial services workforce.
Among more than 210,000 UK financial services employees analysed, the data shows:
- Employees from lower socio-economic backgrounds take around two additional years to reach senior roles.
- Only 38% of employees from lower socio-economic backgrounds report access to AI tools, compared with 48% of those from higher socio-economic backgrounds.
This gap matters because teams building and governing AI systems shape how those systems understand customers.
When leadership teams lack socio-economic diversity, important consumer realities can be missed — increasing the risk that automated systems unintentionally exclude the very groups they are meant to serve.
Everyday barriers could become automated exclusions
The submission highlights several areas where AI-driven automation could unintentionally magnify existing challenges for vulnerable consumers:
- Digital identity checks that fail without reliable devices or connectivity
- Credit scoring models that misinterpret irregular income patterns
- Automated customer support that assumes high digital literacy
- Fraud detection systems that wrongly restrict accounts used in unusual financial circumstances
Individually these may appear as minor friction points. At scale, however, they risk turning small barriers into systemic exclusion.
Recommendations for the FCA
Progress Together is calling on regulators and firms to embed socio-economic considerations into AI governance as part of Consumer Duty responsibilities.
Key recommendations include:
- Testing AI systems for socio-economic bias, including patterns linked to irregular income and financial volatility
- Ensuring transparency and human review, so consumers can challenge automated decisions
- Explicitly considering lower socio-economic consumers in AI risk frameworks
- Strengthening senior leadership accountability under the Senior Managers & Certification Regime (SMCR) for inclusive AI outcomes
A defining moment for financial inclusion
AI will shape how millions of people interact with financial services in the years ahead.
Progress Together’s submission argues that inclusive design and diverse leadership will be critical to ensuring that technological innovation expands financial access rather than narrowing it.
As the FCA considers the long-term future of consumer finance through the Mills Review, the organisation is urging regulators and firms to treat socio-economic inequality as a core conduct and fairness issue in the age of AI.
Further reading
Progress Together’s Rise with AI report explores how workforce inclusion and AI adoption intersect across UK financial services.
Download the report and explore resources on socio-economic diversity at here
