Fashion Tech is becoming increasingly relevant in the race of different sectors to keep up with the progression of artificial intelligence (AI). The innovative developments in the fashion industry incorporating technology and AI into their products have spanned from an umbrella that can warn you when it’s going to rain and wallets able to charge your smartphone, and now to Facebook’s new AI fashion tool: Fashion++.
Fashion++ is a system that uses a network to ‘recognise garments and offer suggestions on what to remove, add or swap.’ It also has the ability to recommend simple changes such as tucking in your shirt or rolling up your sleeves. The system uses a discriminative fashionability classifier that is trained on thousands of publicly available images of outfits that have been judged to be stylish. Researchers then changes an item of these outfits with the least similar alternative, to create what they deem as ‘unfashionable’. The system claims to be original in the sense that previous work has only focused on changing or suggesting an entire outfit, while Fashion++ offers subtle, minimal changes to make your chosen outfit more stylish.
As novel and interesting that this new tool may be, it gives rise to the ever-relevant questions of data protection and security. Although Fashion++ has only been used so far for research purposes, if this system and other similar Fashion Tech products come to fruition, developers and companies ought to ensure the utmost compliance with the relevant data protection laws.
Amongst other things, systems such as Fashion++ would gather photographs of its users and information and health data such as their body shape and size which are classed as sensitive data under the GDPR and therefore require extra protection. How they will treat such sensitive data is one issue, but how they will collect data compliantly poses further questions.
European privacy regulators
The lack of transparency of these technologies is the main issue raised by European privacy regulators. The concern is that customers are not aware of when and how their personal data is collected, the purposes for which their personal data is processed and to whom data are communicated. The problem relates to the level of information and the type of consent that is required by customers for the usage of such technologies especially taking into account that the data collected through them can generate detailed profiles of users.
Many believe that players in the fashion industry should be investing more and more in AI, with the prediction that AI tools, such as Fashion++, will eventually replace human-held roles such as designers and stylists. Start-up EU companies like the UK’s Threads and the German company Outfitterys, which use artificial intelligence to power an online personal styling service, are implementing concepts similar to Fashion++ and have privacy policies that appear transparent and which seem to address all of the primary concerns of AI and GDPR compliance.
AI is seemingly necessary to impress the new generation, but, whilst prevalent in many other fields, it has been slow to make significant contributions in subjective fields such as fashion and retail. As much promise and benefit Fashion++ and similar systems offer, let us hope they are as dedicated to consumer privacy as they are innovation.