By Erin Nicholson, Global Head of Data Protection and Privacy at Thoughtworks
Over half (51%) of European consumers want more personalised financial services, but a
significant minority (22%) are less comfortable sharing data for this purpose compared to last
year, according to a report by Twilio. This highlights the core tension in today’s financial
landscape: personalisation and privacy.
Consumers crave tailored financial advice and products. They want their banks and financial
advisors to understand their unique needs and goals. Yet, data privacy regulations like GDPR and CCPA make leveraging personal data for such purposes a challenge. These regulations restrict how financial institutions can collect, store, and use customer data.
As a data protection and privacy specialist, I am fascinated by bridging this gap. I question how we can achieve personalisation for clients while remaining compliant with these regulations?
The answer lies in a three-pronged approach utilising Artificial Intelligence (AI): leveraging both predictive AI and generative AI (GenAI) and also leveraging Privacy Enhancing Technologies. This approach empowers financial institutions to personalise the client experience while safeguarding sensitive data.
AI-driven lead generation with privacy at its core
Traditional prospecting methods often rely on incomplete data or outdated strategies. Sifting
through vast datasets to identify potential clients can be a time-consuming and inefficient
process. Here’s how AI can help:
Predictive AI can analyse anonymised or aggregated data sets to uncover patterns and trends. This data can be used to create a “probability-weighted list” of potential clients, highlighting those with a higher likelihood of being receptive to specific financial products or services. This approach provides valuable insights without requiring access to sensitive personal information.
Cross-selling reimagined: connecting the dots without data sharing
Cross-selling within a financial institution can be a powerful strategy to deepen client
relationships and drive revenue. However, identifying potential connections between existing
clients and those who might benefit from products offered by different divisions has always been a challenge due to data silos and privacy concerns.
Here’s where GenAI comes in.
GenAI, Federated Learning, and Homomorphic Encryption unlocks the power of graph-based
algorithms. These algorithms can analyse connections between data points without actually
sharing the underlying sensitive data itself. Imagine a system that can identify potential
cross-selling opportunities between different client segments, allowing banks to recommend
relevant products or services while maintaining strict data privacy boundaries.
The power of combining personalisation and privacy
This two-pronged AI approach offers significant benefits for financial institutions:
● Increased efficiency: AI streamlines prospecting efforts, allowing institutions to focus
resources on qualified leads.
● Enhanced customer experience: Personalised recommendations based on anonymised
data insights foster stronger client relationships.
● Reduced regulatory risk: Minimising reliance on sensitive data minimises regulatory risks
associated with data privacy violations.
The broader potential of genAI
GenAI’s potential extends beyond initial client acquisition and cross-selling. Imagine, for example, using genAI to create educational content tailored to each client’s needs and financial literacy level. This empowers investors to make informed decisions based on clear and relevant information, ultimately strengthening the client-advisor relationship.
Responsible AI adoption: a critical priority
While genAI offers exciting possibilities, responsible adoption is crucial to ensure the protection of the public’s data. Here are some key considerations:
● Focus on high-value use cases: Identify genAI applications that deliver significant value
while minimising complexity and cost.
● Ensure data security: Implement robust security measures to safeguard sensitive
customer data from potential risks associated with genAI models.
● Combat bias and factual errors: Be mindful of potential biases in training data and
incorporate human oversight to prevent biased or inaccurate outputs.
● Leverage Privacy Enhancing Technologies: PETs such as Federated Learning and
Homomorphic Encryption will enhance the utility of your data without infringing on
privacy.
By embracing AI in a responsible manner, the financial services industry can achieve its
personalisation goals while ensuring customer data remains protected. This paves the way for a future where personalisation and privacy go hand-in-hand, fostering a more secure and
empowering financial landscape for all.
