AI is helping businesses learn what customers will pay – and workers will accept

Regulators are examining a new aspect of online shopping that uses personal data to determine individual willingness to pay. The Federal Trade Commission is reviewing an enforcement policy for personalized pricing, as sophisticated algorithms could tailor prices and discounts to specific customers. Meanwhile, Consumer NZ warns about the extensive data collected through supermarket loyalty programs, suggesting retailers could gain detailed insights into shopping habits and individual payment thresholds.

This raises concerns about the delicate balance between businesses' ability to understand customers' financial limits and workers' willingness to accept labor. At the University of Auckland Business School, the focus is on how businesses create value and become more efficient, but the same principles apply to workers and customers.

Algorithms are reducing the uncertainty surrounding individual limits, with digital platforms observing thousands of decisions. While there's no strong evidence that major companies currently know the precise financial breaking points of everyone they deal with, algorithm-mediated pay, personalized incentives, discounts, and offers are already a reality.

Companies like Lyft have documented systems that determine which drivers receive incentives, with some earning challenges personalized. Studies on Uber trips in the UK have found that dynamic pricing could lead to lower real hourly earnings and greater inequality. Additionally, the US Federal Trade Commission discovered that pricing intermediaries have access to information such as location, demographics, browsing histories, and even mouse movements, influencing prices, discounts, and promotions.

Retailers don't need to set different prices for each customer but can offer discounts to those predicted to walk away and withhold them from those predicted to buy. Traditionally, there's been uncertainty on both sides – employers know more about wage structures than workers, and sellers more about margins than buyers. However, algorithms now risk reducing this uncertainty in favor of firms.

While AI-driven personalization can have genuine benefits, such as improved matching, targeted discounts, and better forecasting, the key issue is how these gains are distributed. The real question is who benefits, whether it is fair, and what happens if all businesses adopt these practices.