Banks Adopt AI Clones for Product Testing

The traditional method of testing new credit card or banking products involved months of regulatory vetting and customer recruitment. However, financial institutions are now adopting a new approach by building artificial intelligence-generated customer clones, also known as synthetic profiles. These profiles cost almost nothing and carry none of the compliance exposure tied to real customer data.

What does this mean for the banking industry? Can synthetic data really replace human customers in product testing? The answer is complex, but the trend is clear: major institutions on both sides of the Atlantic are already using synthetic data ecosystems to train AI models and test new products.

How Synthetic Data Works

U.S. Bank, for example, deploys synthetic audiences to model consumer segments such as high-net-worth households, testing messaging and refining campaigns before launch. JPMorgan Chase generates synthetic financial data to simulate market behaviors for risk management and product design. NatWest, Monzo, and Santander are also using synthetic data ecosystems to train AI models.

But how does this process work? Synthetic data is created using advanced algorithms that mimic real customer behavior. This data can be used to test new products, refine marketing campaigns, and even simulate market behaviors. The benefits are clear: synthetic data is faster, cheaper, and more efficient than traditional customer testing methods.

However, as Mudit Gupta, EY’s AI practice leader for Americas financial services consulting, notes, governance is a critical issue in the adoption of synthetic data. Most banking leaders believe that agentic AI can move faster if governance were not perceived as a constraint. But in practice, governance is what makes these systems deployable at scale.

Regulatory Frameworks and Governance

In the U.K., the Financial Conduct Authority (FCA) has launched its AI Live Testing initiative, which aims to bring the practice of using synthetic data inside a regulatory framework. The initiative includes major institutions such as NatWest, Monzo, and Santander, and will test the use of synthetic data in areas such as agentic payments, anti-money laundering detection, and know-your-customer checks.

The FCA’s initiative is seen as a positive step towards regulating the use of synthetic data in the financial sector. However, as Gupta notes, synthetic data is often treated as inherently safe, which is not the case. It can leak sensitive signals through inference and linkage risks, and can also replicate and scale historical biases, embedding them behind a layer of abstraction that makes them harder to detect, audit, and challenge.

So, what are the implications of synthetic data for regulatory frameworks? The answer is complex, but it is clear that regulators will need to take a closer look at the use of synthetic data in the financial sector. The FCA’s initiative is a step in the right direction, but more needs to be done to ensure that the use of synthetic data is transparent, accountable, and fair.

Adoption and Implementation

The scale of adoption of synthetic data is already significant, with major institutions on both sides of the Atlantic using synthetic data ecosystems to train AI models and test new products. The technology is also moving into treasury and finance operations, where forecasting models have historically relied on data that quickly becomes stale.

But what are the challenges of implementing synthetic data in the financial sector? The answer is complex, but some of the key challenges include ensuring the quality and accuracy of synthetic data, addressing governance and regulatory issues, and ensuring that the use of synthetic data is transparent and accountable.

Despite these challenges, the benefits of synthetic data are clear. Synthetic data can help financial institutions to test new products faster, refine marketing campaigns, and simulate market behaviors. It can also help to reduce the risk of non-compliance and improve the overall efficiency of the product testing process.

What Should You Do About Synthetic Data?

So, what should you do about synthetic data? The answer is simple: stay informed, stay vigilant, and ensure that you understand the implications of synthetic data for your business. Whether you are a financial institution, a regulator, or a consumer, it is essential to understand the benefits and challenges of synthetic data and to be prepared for the changes that it will bring.

As the use of synthetic data becomes more widespread, it is likely that we will see significant changes in the financial sector. From the way that products are tested and launched, to the way that regulatory frameworks are designed and implemented, synthetic data is likely to have a major impact. So, stay ahead of the curve, and ensure that you are prepared for the future of financial services.

The use of synthetic data is a game-changer for the financial sector, but it also raises important questions about governance, regulation, and transparency. As the technology continues to evolve, it is essential that we prioritize these issues and ensure that the use of synthetic data is fair, accountable, and transparent.

What does the future hold for synthetic data? The answer is complex, but one thing is clear: the use of synthetic data will continue to grow and evolve in the coming years. As the technology improves, we can expect to see more widespread adoption, and more significant changes in the financial sector. So, stay informed, stay vigilant, and ensure that you are prepared for the future of financial services.

Synthetic Data in Financial Services – Disclaimer

This article does not replace professional advice. Outcomes may vary by individual circumstances. Consult a qualified financial advisor or regulator for specific guidance on synthetic data in financial services.