Data deficit is slowing down use of AI in consumer enforcement
The EnfTech results are in: mixed reviews for AI?
On 29th Feb, UNCTAD in association with Poland’s Office of Competition and Consumer Protection (UOKiK) organised a webinar on AI and consumer protection: harnessing tech to better enforce consumer law.
It featured talks from UOKiK and collaborators on their project on the use of AI to detect dark patterns. Prof. Christine Riefa presented the results of our EnfTech report focussing on our findings concerning the use of AI in enforcement.
A trend is clear - AI adoption is spreading in consumer agencies that are more advanced in their EnfTech journeys. However, AI use remains behind the 'state-of-the-art' curve compared to use cases further afield.
This is in part explained by the availability of data other agencies may have compared to consumer agencies and also because use cases outside the consumer field had different objectives explaining why AI may be favoured as a tool. Consumer agencies are also not yet well resourced to develop and roll out AI solutions.
Rolling out EnfTech and more specifically AI requires working through a series of problems, explored in our report.
While AI is not always the tool of choice and gains can be made at any stage of the EnfTech journey, the use of AI is showing promise to perform time intensive tasks. But it can only work if there is good data and a lot of it to train the AI on and use as a reference point.