Facial Recognition AI will use your facial expressions to judge creditworthiness
“Sorry, Mr. Jones, we can see from your face you’re not likely to repay a loan.”
Credit institutions are poised to use a combination of artificial intelligence and facial recognition to instantly read the facial expressions of applicants to assess their likelihood of loan repayment. Sound far-fetched? It isn’t. It’s already being done.
The South China Morning Post reports that Ping An Puhui, a Chinese micro lending unit of China’s second-largest life insurer, has developed a digitalized loan process that can “analyse facial expressions of applicants to determine their willingness to repay the loans.” The company contends that as a result of using new technologies, including facial recognition and big data, it has seen its customers “more than doubling to 5.5 million from 2 million a year ago,” and its loan default rate drop, without the necessity of expanding its staffing.
Facial recognition for identity verification and mobile payments just got a big boost with the latest iPhone launch. Thanks to Apple, consumers will become far more accustomed to the use of facial recognition for identity verification and digital payments. Apple’s iPhone “Face ID” is used not just to unlock the phone, but also to make Apple Pay purchases. How much more will consumers accept in AI and facial recognition technologies?
One peer-reviewed study already asserts that facial recognition can be used to detect sexual orientation. Is there anything to stop banks or others from turning down a loan based on their reliance of AI to make an assessment from a facial recognition analysis?
In the US, consumer safeguards from the Fair Credit Reporting Act (FCRA), enforced by the Federal Trade Commission (FTC), would likely be triggered by the use of AI and Facial Recognition to deny a loan. FCRA, a federal law, requires anyone who uses a “credit report” to deny an application for “credit, insurance, or employment — or to take another adverse action against you” to provide a consumer with the information that was used to make an adverse decision. While the law does not speak to the use of facial recognition technologies, it is a law that is generally broadly construed to optimize its consumer protections of “fair and accurate” credit reporting. If facial recognition and AI are used to deny a loan or reject an applicant as an employee, there is a good likelihood that the FTC would conclude that FCRA protections would apply.
As Facial Recognition and AI move into the mass market, enterprise, government, and consumers should carefully assess how the combination can and should be used. Weaving facial recognition with AI promises great efficiencies that could benefit government, business, and consumers. To achieve the fullest benefits of both technologies, however, the highest standards of ongoing inquiry and auditing of core processing should be insisted upon to guard against inaccuracies and bias. Too much rests upon “digitalized” processing not demand the highest standards in every respect.