The future of fashion marketing: is AI giving the industry a makeover?

Whether it’s dressing the Pope in a Balenciaga puffer jacket, or speeding the process of crafting next season’s collection, artificial intelligence is expanding the scope of the industry.

Just like in other sectors, generative AI has the potential to help fashion businesses become more productive, get to market faster, and personalise the customer experience.

According to a study by McKinsey, generative AI could add up to £216 billion to the apparel, fashion, and luxury sectors’ operating profits.

However, change tends to be accompanied by disruption. As generative AI becomes more intuitive, creatives and models fear the extent to which AI will become a replacement rather than a companion of their work.

How is AI changing the way the fashion industry works?

To understand how AI is giving the fashion industry a glow up, we spoke with Startups 100 alumni, Intelistyle. The startup uses artificial intelligence to bring personalised, in-depth style advice to customers of some of the world’s most popular brands, including Tommy Hilfiger and Max Mara.

One exciting frontier AI is crossing is outfit generation. Kostas Koukoravas, Co-Founder of Intelistyle, explains, “What we do is crawl the web, understand what looks good, train an AI model to be stylish, and then help customers see entire outfits.”

In addition to enhancing the customer experience by giving every buyer their own personal stylist, it also answers to a growing consumer expectation: personalisation. 80% of consumers are more likely to buy from a company that provides a tailored experience, while 66% of consumers expect brands to understand their individual needs.

However, fashion brands will often experience overdemand for certain trendy pieces and run out of stock. Koukoravas notes that AI can be trained to identify those pieces in iPhoto or on social media and then find the closest match that is in stock.

What could be considered more important is the way AI is helping fashion become inclusive. Historically, the fashion industry has been criticised for giving the spotlight to the classic model archetype – tall, slender, with knife-sharp cheekbones. As a result, many feel underrepresented or frustrated when a certain garment doesn’t fit or look the way it does on a marketing ad.

“What really excites us, and something that we’re working on, is the ability to visualise how a particular garment might look on different models as that might be a better representation of models with different body types and different skin tones,” explains Koukoravas. “I think it makes fashion a lot more inclusive and it makes sense as it is relatable to customers.”

Could models go out of fashion?

If you remember the Sony World Photography award being snapped up by an AI generated image, you wouldn’t be alone in questioning whether real models still have a place in ads and magazines.

From the perspective of brands, replacing models with AI certainly has its allure. Paying professional models for a photoshoot can be expensive – especially if you go directly to a modelling agency.

“I think it enables retailers to do more because if you think of a retailer and all their costs, the pressure on gross margins, it is a tough industry and they don’t have the ability to do multiple photoshoots,” points out Koukoravas. “AI allows them to create 10 different versions of a photoshoot that would not have been done anyways, because the cost would’ve been too high.”

Nevertheless, physical models continue to be the inspiration for AI, which means models still have a stable catwalk to walk on. As regulation governing AI is still drafted, Koukoravas notes models can still raise intellectual property as a defence to claim their place in the spotlight.

Photoshoots using generative AI would still theoretically need the approval of models or their agencies as it is still their image appearing on the ads – even if they didn’t pose in front of the camera for that particular ad.

“We, the artists, are saying AI is kind of copying our work because it can learn what we do and then just replicate it, and that’s a difficult question on where the line is,” says Koukoravas. “But AI is a great tool and you certainly don’t want to stop its progress with some really regressive intellectual property laws.”

If anything, the leading role of creatives and models in the fashion industry is still safe from a technological takeover. Generative AI is trained based on the data it is fed, which is still protected by creative licences.

The outcome, therefore, might be that AI will be used as a tool to speed up the creative process and sprout inspiration, rather than a replacement for human creativity.

Will the highstreet still be part of the runway?

Efficient personalisation and a smoother online experience as well as the rise of the fashion circular economy could spell an uncertain future for the highstreet. According to a study by SAVOO, 25% of consumers are spending on the high street less than once a month.

However, Koukoravas doesn’t think that the big fashion stores will be closing their businesses. In fact, AI could play a key role in revitalising the high street shopping experience.

“I don’t think that it is going to significantly impact physical stores. I just think it’s going to make online shopping smooth and easier for people, and a more pleasant experience,” predict Koukoravas.

Regulating the catwalk

However, if AI is to thrive in fashion, regulation needs to be designed to foster innovation while still upholding ethics at its regulatory core. Although the government’s AI whitepaper is a first step, fashion startups, like Intelistyle, are still hoping for a clearer idea on the limits.

“I think it’s an iterative process and a good thing that the government is thinking about it, but everyone needs to engage a lot more and try and put some meat around it to get a concrete set of regulations.”

Startups recently spoke with an AI Startups 100 Alumni, Rafie Faruq, the CEO of Genie.ai, an AI-based legal assistant. During this interview, it was surprising to note that startups are not included in the policy making process.

Faruq attended an All-Party Parliamentary Group meeting, noting “There were, I think, no startups there and pretty much no machine learning engineers or academics. It was all policymakers and political scientists, which was a bit worrying to me,” confesses Faruq.

“I think you need to hear the voice of people who are on the cutting edge of creating new products, new services, and then you need that pool to be diverse in terms of socio-economic background, gender diversity, and every other form.”

Is AI taking over the fashion industry?

AI is not yet ready to become a replacement for creatives. In fact, the picture we used in this article was generated by AI and, if you look closely, the blonde model on the runway sporting an orange skirt is missing a leg.

“Creators need to be there because creators are the lifeblood of everything,” emphasises Koukoravas. “AI effectively copies or learns from that creativity, and I don’t think that AI will ever be able to do as good as creators.”

Generative AI still needs to be trained on human creativity. If anything, some processes in the fashion industry will be automated, but rather used as a helping hand rather than a replacement for a human. Importantly, AI can also help digitise and speed up creative processes, which means prototypes can be created more efficiently and produce less waste.

Although AI can help transform someone’s look or find them the perfect outfit, it’s unlikely there will be an automated replacement for a modern-day Coco Chanel. Creativity is something that is an innately human trait.