Finance

AI-Driven Financial Inclusion: Unlocking Banking and Investment for Underserved Communities

By Livia Bernardini , CEO of Future Platforms

Financial inclusion is critical to economic well being. Having access to affordable financial products and services allows individuals to build wealth, manage money and invest in different assets, from businesses to government bonds. This in turn increases tax revenues. The importance of financial inclusion is such that it is seen as a driving force for seven of the United Nation’s 17 Sustainable Development Goals.

Yet a significant proportion of the world’s population remains underserved by financial products. The World Bank estimates that, even with the expansion of digital financial services, 1.4 billion adults, or 24% of the global population, do not have access to a transaction account, a foundational part of financial inclusion.

The causes for this lack of inclusion vary. Some segments are considered too high a risk for traditional providers, while a lack of access to relevant products hits others, as does financial illiteracy. Individuals in low-income countries are more likely to be unbanked, as are micro, small, and medium-sized enterprises (SMEs). This can have a detrimental effect on productivity, long-term growth projections, and job creation.

This isn’t a new problem; since 2010, more than 60 countries have either launched or developed national financial inclusion strategies, bringing together stakeholders including regulators, private sector organisations, and government bodies in telecoms, agriculture, and education. Technology has long been seen as a key enabler, with innovations from mobile banking to digital identification helping to close the financial inclusion gap.

AI’s Role in Financial Inclusion

No innovation promises as much as AI. Combining data, large language models, and advanced analytics, AI tools offer a significant opportunity to bridge the gaps that traditional finance struggles to address. These include:

  • Transforming access to banking and investments: AI tools can make financial services more available to people without the credit history or income that traditional providers look for.

For instance, Acorns uses AI to automate savings, round up purchases for micro-investments, and personalise budgeting tools, all of which adapt to user behaviour so that all suggestions remain tailored to individual customers.

  • Credit scoring innovation: With its ability to process and analyse larger volumes of data at a greater pace, AI can be deployed to use alternative sources to generate credit scores, allowing underserved audiences to access products previously unavailable via traditional scoring methods. It can also help providers develop new services that keep risk at an acceptable level.

 It’s an approach Experian and Credit Karma are using to broaden financial access. Advanced modelling allows providers to more accurately assess credit risk, identify potential fraud, and understand customer behaviour, which in turn informs how they price loans and accept applications. Elsewhere, Tala’s personalised credit platform not only underwrites customers who have never interacted with traditional banks, but does so in seconds, accelerating decision-making and improving acceptance rates. 

  • Improving financial literacy: Customer-facing applications can be used to improve financial literacy among underserved communities, with personalised conversations, jargon-free responses, and even via multiple languages and dialects.

Cleo provides customers with a way of building credit and better financial habits. Positioning itself as the world’s first AI Money Pro, it uses an accessible, informal tone of voice to engage with customers.

Using AI to Improve Financial Inclusion Ethically

Of course, there are ethical issues that must be considered when deploying AI. These include data privacy, overcoming biases in models and algorithms, and transparency.

AI doesn’t exist without data, but that does not mean it should have access to all data. Companies that adhere to regulations without hampering their innovation capabilities are the ones that are targeted in the data they feed into their models, minimising the amount of user data required. They are also clear on what data they are using and how it benefits the user; Branch, one of the world’s most popular finance apps, is open and upfront with its customers that it requires their smartphone data to determine loan eligibility in seconds.

Tackling bias in algorithms is an ongoing challenge; ultimately, everyone has unconscious prejudices, and over time, that has infiltrated much of the data AI models use. Improving diversity across organisations, from design and development to customer experience and product marketing, is vital if we’re to dramatically reduce the impact of bias becoming embedded into services.  Regular testing and analysis of model assumptions help catch anything that enters production, while open-source bias mitigation tools can measure and rectify issues in data sets. It’s quite simple: if we don’t tackle bias, then it becomes ingrained, and that could ultimately undermine the whole point of using AI to improve financial inclusion.

Transparency is essential for building consumer trust in AI. However, many AI models operate as black-box systems, offering little insight into the data used and the decision-making processes behind outputs. Nearly three-quarters of consumers say they are concerned about certain AI technologies, highlighting the importance of being transparent and clear as to what has been used and why.

It’s important to remember that the average customer is not going to have an in-depth understanding of AI. As such, those tools that are open about how data is used, how it informs outputs, and, most crucially, what that means for end users will be more likely to be trusted by their customers.

Bridging the AI Gap to Improve Financial Inclusion

As well as ethical considerations, there are other obstacles to overcome if AI is to help close the financial inclusion gap:

  1. The Execution Gap – AI is great at planning and theory, but its execution still relies heavily on human intervention. With AI skills in short supply, this can hamper scaling AI from pilots to production.
  2. The Learning Gap – While there are a huge number of AI tools available, many of them are generic large language models. They can struggle to retain long-term context or deliver outputs relevant to specific use cases, requiring further human intervention to meet business needs.
  3. The Coordination Gap – Interoperability has not been at the forefront of AI tool development; many of them function in isolation, unable to collaborate effectively across systems or platforms.

Does Agentic AI Offer a Solution?

Agentic AI could be the answer to AI’s deployment issues. It can retain and build on contextual information, coordinate with other systems, and act autonomously, within predefined ethical and procedural boundaries. Specific agents can be developed to meet the exact needs of a business or its use cases.

This means more reliable, adaptive, and integrated solutions that respond dynamically to user needs. When it comes to financial inclusion, these AI agents could reduce time and increase access, thanks to enhanced user experiences, automated decision-making, and reducing the need for constant human oversight, while still maintaining trust and explainability.

What’s critical is that businesses deploying agentic AI do so to support their purpose, not just to solve an efficiency challenge. That comes back to the point about having specific use cases, and not just a blanket approach.

Towards an Inclusive Financial Future

From real-time credit scoring drawing on non-traditional sources, to personalised financial literacy training and accounts that learn from customer behaviours, AI has the potential to drive true financial inclusivity in underserved communities.

The deployment of AI is not without its challenges; many organisations do not have access to the skills needed to turn generic tools into truly transformative systems. Agentic AI offers a solution, but any use must address data privacy, bias, and transparency concerns while remaining true to the organisation’s overall purpose.

Yet the organisations that get the deployment of AI right will be at the forefront of delivering new economic opportunities to a significant proportion of the world’s population.

Author

Related Articles

Back to top button