The Retailer Spring Edition 2023

SPRING 2023

19

4. The balance of AI with real data Consequently, developers are creating systems which bypass personal data by claiming it is not possible to identify individuals. However, this does not go far enough to conclude that no personal data is being collected and processed if an individual simply cannot be named. These systems also carry the risk of reputational pushback from privacy campaigners. From the retail sector’s perspective, not being able to identify customers after their in-store experience results in little or no promotional follow-up nor longer-term behaviour analysis or customisation.

5. The regulation of AI Despite AI technology continuing to develop at a rapid pace. the regulation of AI also continues to develop. Several drafts of Europe’s future AI Act considers that where systems are “at a distance”, and happen without the “active involvement” of participants, it will still be perceived to be a “high risk” system. Consequently such systems will still necessitate public authority notification and public consent in order to be compliant. Lessons can also be learnt from the lack of information provided to customers relating to existing biometric tracking experiments. Several of these have fallen foul of the law in the US, on the grounds that clear prior information was not given to customers entering stores.

As with the presentation of cookie banners, a delicate balance will exist between provid ing sufficient and excess information. Both approaches risk alienating customers at an opportune moment of engagement. In order to overcome the increased regulation of AI, whilst deploying innovative and insightful technologies, retailers will instead have to focus on their brand winning the hearts and minds of consumers for successful deployment.

Shruti Goel shruti.goel@shoosmiths.co.uk

Alice Wallbank alice.wallbank@shoosmiths.co.uk