Fintech Insights

Keeping Machine Learning, Ai Ethical for Customers

February 26, 2018

“It’s really important that we have more intelligence than artificial.” -- Apple co-founder Steve Wozniak’s thought on artificial intelligence (AI) at Money20/20

Financial institutions can use machine learning applications to build a data strategy that becomes smarter and more robust over time. When used correctly, AI can reduce payment fraud, personalize customer experience, boost the relevance and profitability of rewards and loyalty offers, and determine credit worthiness and risk. Yet Maria Schuld, FIS Group Executive, Financial Services says banks who use AI and machine learning must be certain to implement ethical approaches to safeguard private and highly personal customer data, and deliver a positive customer experience.

Banks must ensure their machine learning strategy is ethical, for these critical reasons:

Precision is paramount

Financial institutions (FI’s) who use AI and machine learning can build models to detect customer behavior patterns, like where and when they shop, how much they spend and how they want to pay. Yet, precision is key to intelligent fraud prevention.

Consider the impact incorrect AI-generated false positives can have on customer experience, retention, and interchange revenue, based on the results of one study by Javelin. According to the study, 15 percent of cardholders surveyed said they’d had at least one legitimate transaction declined in the past year. Of those, nearly 40 percent abandoned use of their card completely after the inaccurate fraud detection.

Credit risk measures must be compliant

AI can help consumers with little or no recent credit histories gain access to credit with alternative data, like telecom or utility payments. AI-based lending platforms can use hundreds of data points to determine creditworthiness, propensity to default on loans and the possibility of fraud.

Yet, financial institutions must assess risk factors associated with AI, particularly in the areas of credit, debit and fraud. They must ensure that data used in the alternative algorithms to judge creditworthiness is accurate, and verify that machine decisions based on alternative data fall within compliance guidelines. Further, the data cannot be correlated with factors that could lead bankers to make illegal or discriminatory decisions.

Financial Institutions can ensure that AI used in credit risk is ethical by:

  • Working with the processor to mine the consumer data that will be used in decision-making to ensure accuracy;
  • Using caution when creating the if/then analyses to confirm if/then assumptions are fair to consumers; and
  • Verifying that criterion used to determine whether consumers are approved fall within compliance guidelines; decisions must be explainable and justifiable.

FIS leverages data from issuers to enable machine learning in the fight against fraud. This empowers financial institutions to leverage the many benefits of AI, while maintaining the controls and parameters necessary to delivering accurate and consumer-friendly models.

Read more about FIS’ view on the deployment of artificial intelligence in banking and questions surrounding the ethical use of data in this article.

Tags: Digital, Innovative Technology, Risk & Compliance