‘Fairness-as-a-service’ helps businesses reduce algorithmic biases

The technology tackles disparities that continue to adversely affect certain groups

Spotted: Traditionally, to obtain a loan, a customer would need to interact with a loan officer, whose discretion could have a significant impact on their access to credit. Analysis by the World Bank has shown that racial proximity can influence that discretion, while another study found that African-Americans and Latinos are far more likely to be denied conventional mortgages than whites. Women too face similar disparities.

Today, algorithmic software is increasingly deciding the outcome of major life events, from surgery to job interviews, mortgage loans, and much more. This brings a number of benefits, yet bias remains rampant in decisions made by algorithms. Now, FairPlay is taking on this challenge. The LA-based company provides the world’s first ‘fairness-as-a-service’ solution to tackle algorithmic bias in decision-making.

With four products, FairPlay helps financial institutions better analyse their customers and improve the equality of decision-making without compromising on risk. For loan providers, the platform identifies good applicants who may have been denied by conventional lending models, flagging them for a second look. The company has also created ‘Fairness Maps’ that visualise the fairness of mortgage approvals for Black, Hispanic, female, and other protected applicants in every county of the US. 

Having recently raised $10 million (around €9.8 million) in Series A funding, the company plans to expand its products for use in fraud, marketing, and insurance, while growing its data science and engineering teams.  

Other recent innovations in diversity and inclusion spotted by Springwise include a ‘spellchecker’ for inclusive writing, software that automatically analyses the diversity of media content, and an artificial intelligence platform that helps companies track areas where they face specific challenges with diversity and inclusion.