The challenge
Mobile phone users are at risk from deceptive patterns
A deceptive pattern is a web, app or UI design that works against you. It might fool you into providing your data or to make it hard to leave a subscription. Deceptive patterns on user interfaces (UI’s) range from the merely annoying, to the malicious.
Tech users, developers, and regulators are all asking: how can we protect ourselves and others from harmful deceptive patterns?
Currently, there are no established tools that help protect mobile phone users.
Our response
Enter UIGuard: a deceptive pattern detection tool
We conducted a systematic analysis of deceptive patterns on mobile UIs to underpin the development of the first data-driven deceptive pattern detection tool: UIGuard.
Using computer vision and natural language pattern matching techniques, we distilled patterns from mobile phone user interfaces to create a large-scale pattern bank, used to identify dark patterns in mobile UI’s.
The results
Users of UIGuard indentify a large range of deceptive patterns
During testing, we found that pre-UIGuard, mobile users could find around 165 deceptive patterns in a sample. But when assisted by UIGuard, this increased to 453. Users were also able to identify new deceptive patterns on their own.
As the first tool to be able detect deceptive patterns in mobile applications, the work has the potential to benefit a variety of stakeholders, including end-users, app providers (designers and developers), and regulators.
We are refining the tool to be able to recognise a wider range of deceptive patterns. Further work will investigate the ability to block malicious designs, rather than just detect them to protect users from harm. The tool will be released to the public next year.