Using AI for consumer protection: creating AI-based personas for mystery shopping

In the age of big data, personalisation of the user’s experience is the norm. The personalisation of content is beneficial to consumers when it helps them to quickly find products and services they are looking for. However, profiling-based personalisation facilitates manipulative techniques that are often discriminatory and infringe on the rights of consumers.

Surfing the internet under a fake identity (an AI-based persona) would allow researchers and consumer authorities to discover online personalisation techniques that use prohibited criteria for targeting consumers. It may also reveal techniques that decrease the ability of consumers to make informed decisions and therefore should be prohibited. However, the question is: How can you develop efficient artificial intelligence techniques to 'trick the trackers' while respecting all legal and ethical rules?

This project involves a feasibility study for an AI-based tool that automatically generates personas and carries out online mystery shopping. The research has academic value as well as potential for creating value for policy developers and law enforcement in the areas of consumer protection and non-discrimination law.