Forget Pinterest boards or resale deep dives — for Gen Z, fashion advice in 2025 is increasingly AI-assisted. A scroll through TikTok reveals a new genre of content: users prompting ChatGPT to build capsule wardrobes, suggest seasonal colour palettes or even guide them through their style identities.
“What do I wear to a second date at a rooftop bar if it might rain?” one user asked ChatGPT in a video posted to TikTok in February, now with 50,000 views. “How do I look professional for a job interview without giving ‘corporate drone’?” asked another, in a video with 30,000 views. The latter is advised by the artificial intelligence to wear a structured blazer in an unexpected fabric (like velvet or linen) or a unique neckline, both of which “can signal personality without sacrificing professionalism”. There’s scores of similar videos, as Gen Zs increasingly adopt generative AI to simplify their lives.
OpenAI is leaning in. In April, in response to a surge in searches on the platforms (around one billion per week) ChatGPT rolled out a new shopping tool. Users can now receive personalised product recommendations with images, aggregated reviews and direct buy links — all via natural language chat. “Instead of juggling tabs or scrolling through endless results, you can just have a conversation,” says Saguna Goel, product lead on the ChatGPT search team, who notes that she’s seeing users get really creative with ChatGPT for fashion advice — for instance, uploading photos of items they already own and generating mood boards and shopping lists based off of it.
OpenAI has announced updates to the shopping experience on its generative AI platform ChatGPT, recommending products and where to buy them. Is this the future of online shopping?
“What’s really exciting is that this movement is being led by shoppers themselves. Gen Z are definitely out front, but people of all ages are getting involved — and stylists and influencers are now following their lead. It’s a fascinating shift, and it’s only just beginning,” she adds. For example, online fashion publication Sheer Luxe recently posted a video of their fashion editors road-testing ChatGPT’s styling recommendations. As did fashion and lifestyle influencer Sheroy.
But not everyone is convinced this is a good thing. “I kind of roll my eyes, I can’t lie — it’s like these days we need everything to be fast tracked and perfectly packaged up for us,” says Macy Eleni, content creator and author of Second Chances: The Ultimate Guide to Thrifting. “As someone who firmly believes that the key to honing personal style is through simply living your life, I feel tools like this while yes, efficient, can take away from the beautiful mess that is figuring out what you truly like to wear.”
If TikTok’s algorithm is flattening everyone into the same five aesthetics, then generative AI might just be the wrecking ball to individuality altogether. “We are already at the collapse of originality,” says Brenda Weischer, influencer and host of ‘Brendawareness’ podcast. “I think we’re all thinking about ourselves too much. My dressing comes intuitively so outsourcing actually seems like such a hassle.”
Generative tools, including ChatGPT, also carry a steep environmental cost, consuming significant amounts of energy and water to operate. Those concerns are compounded by the way such tools streamline the shopping experience, potentially fuelling the very habits the industry claims it wants to curb.
Still, considering how tiresome and complicated online shopping can be, it’s easy to see the appeal — especially for those with only a passing (if any) interest in fashion and need a quick recommendation of what to wear to a friend’s wedding. So how can brands show up meaningfully in a space where large language models are shaping purchases?
OpenAI announced new shopping features on ChatGPT that offer conversation-based recommendations that are, for now, free of ads. Is this the next era of e-commerce?
How brands are tapping in
For the past decade, digital marketing has been a relentless game of trend-chasing: hop on the latest micro-aesthetic, get an influencer into your product, catch the TikTok algorithm at just the right time. But as more consumers start asking chatbots what to wear, and receive tailored style advice and shopping recommendations in return, the battleground is shifting. Instead of optimising for virality, brands are now figuring out how they can show up in the chat.
.jpg)
For those who get it right, the potential is enormous. There’s currently no paid placement on ChatGPT. So AI recommendations flatten the playing field. “As a digitally native brand, we recognise the importance of being discoverable where today’s customers are searching,” says Leen Abdelnour, founder of jewellery label Lynyer. “We’re optimising our product data across platforms like Google Shopping and Shopify, which we understand are often indexed by AI models.”
For small businesses, it’s a particularly promising shift. With ChatGPT’s latest shopping feature offering product images, reviews and direct links — without companies able to buy ads to boost their listings — brands like home fragrance brand Kōdō London see a chance to compete on clarity and relevance, not budget. “The new features of the shopping assistant mean that as a small business we won’t have to compete with advertising budgets,” says Kōdō founder Yasmin Khalil. “Consumers who align with our brand can find us purely through AI recommendations.”
So how can brands show up in these spaces meaningfully? “ChatGPT interprets the intent behind the request based on the question and any other available context. ChatGPT will consider general factors, such as price, customer ratings and information on web sources about [the desired product], as well as specific criteria provided by the user, like sizing and any desired styling,” says Goel. “It also considers responses generated by ChatGPT before it provides the search results and of course our safety standards.”
For brands, this means “a higher demand for strategic clarity and positioning services to support multi-layered visibility; think consistent language, distinctive tones and culturally resonant moments,” says Holly Brunskill, managing partner at B The Communications Agency. “On top of this, as ever, brands would be wise to seek allies and relationships with industry figureheads and editorial critics who enduringly offer trend legitimacy and original storytelling opportunities to support their narratives.”
It also depends on the questions asked by the user. Say, for instance, I ask: “I want to buy a dress for a wedding that’s under £150 and will go with my new red heels. Can you pull up a few suggestions?” Some of these factors will be more relevant than others; in this instance, price and styling with red shoes were specified and would be prioritised, says Goel.
Curious myself, I ask ChatGPT this exact question. It returns a tiered maxi dress from Next (£80), a floral Ted Baker dress currently listed for £50 on Ebay, an Asos pleated cami midi dress (£12) and a red Fashion Nova corset-waist satin dress (£32). For comparison, when competitor chatbot Claude suggests an H&M satin pink dress (£50), a Boohoo plus-size wrap-over dress (£35) and a New Look wrap-front dress (£40), it tells me to check each site online for availability, as it isn’t able to provide specific links. Meanwhile, Google’s Gemini responds with general fashion advice around colour palette, style and where to look, rather than specific product recommendations — each demonstrating that ChatGPT’s results are noticeably more precise and actionable at this stage.
When I ask ChatGPT why it chose those particular styles, it responds: “I chose these styles based on a few key criteria that align with your original brief — a wedding guest dress under £150 that pairs well with red heels — while also factoring in versatility, formality and visual harmony.” It then goes on to justify why the selections complement the colour red, are wedding appropriate and flattering, as well as the price point.
Noting that, apart from the Ebay listing, most of these options aren’t sustainable, I ask specifically for eco-friendly picks. It then recommends a 100 per cent organic cotton option from Thought Laelia, a £60 dress from Seba made from eco-friendly Tencel, but — somewhat unexpectedly (and wrongly) — also a £14 green floral tea dress from Asda. Such errors, our research has found, are because ChatGPT sometimes needs very detailed prompts to deliver accurate results for sustainable fashion, which the average consumer may not think about.
ChatGPT’s new shopping recommendation features could be a breakthrough moment for conscious consumption — but the risk of greenwashing is high.
Shelley Cook, co-founder of activewear brand Another Version, shares this concern when it comes to inaccuracies if users don’t ask the right questions. “An outfit designed for leisure might be recommended for high-intensity workouts, which could lead to customer dissatisfaction,” she says. “We’re also aware of the challenges around attribution and want to ensure our designs are credited and linked correctly. However, we view these as manageable risks — and we’re committed to proactive collaboration with platforms to help shape how our brand is experienced in AI environments.”
OpenAI is aware of the stakes. “It’s still very early days for shopping in ChatGPT, and we’ll continue to bring merchants and brands with us as we quickly learn and iterate,” says Goel. “One of the things we’re exploring is an easy way for merchants to provide product feeds directly to ChatGPT, helping ensure more accurate, up-to-date listings.”
‘Personal Style’, brought to you by AI
The cultural conversation has been dominated by the idea that social media algorithms are homogenising everything from music taste and movie preferences to personal style. Unlike the mass broadcast of viral TikToks that inspire everyone to wear the same thing, AI interfaces respond directly to you. They ask follow-up questions, entertain your specific rabbit holes and speak in your chosen tone. Outside of Europe, ChatGPT will also memorise your searches to get to know you better and provide better recommendations over time.
But that intimacy, some argue, is an illusion. “The dynamics that algorithmic recommendations have started will only be accelerated by generative AI, whether it’s styling or search,” says Kyle Chayka, author of Filterworld: How Algorithms Flattened Culture. “On social media feeds, algorithms push us towards a generic result; generative AI creates that generic result immediately because all it can do is produce results from the data that’s fed into it.”
ChatGPT can recommend a look, a brand, a vibe — but it’s pulling from the same pool of reference points everyone else is using. The more people rely on these tools to shape their style, the more that style converges, some fear. Moreover, Chayka notes that it’s more of a “pseudo-personalisation”, noting that the thrill of building a look over time — discovering an obscure brand on a blog, trying something outside your comfort zone, growing into it — is being outsourced to an interface.
“You make mistakes. You experiment. You stumble on something unexpected. Magic is born through the mess. That’s how style becomes yours,” says Eleni, who has built her platform around secondhand style, while her book offers a manifesto for slow, intuitive dressing. “But if people already used to instant gratification — the ones hooked on viral hauls — start turning to AI for fashion advice, I worry we’ll just see more of the same: more overconsumption, more fast fashion, fewer meaningful choices.”
Many people also now want to learn about what they’re wearing, not just be told what to wear. While AI tools like ChatGPT offer quick styling suggestions, they lack the nuance, context and lived experience that come from a human perspective. Content creator and vintage enthusiast Tanya Ravichandran notes that most people follow her to learn about the history of fashion or to see a campaign that’s visually interesting or educational. “AI might offer convenience, but it lacks that human filter, the depth of passion or experience that gives context to a recommendation. That still really matters to some,” she explains.
Ravichandran is also sceptical that generative AI can replicate the kind of relationship people form with their favourite style influencers. “That archetype craves the personal element — the obsession, the intimacy, the ritual of following someone’s life and aesthetic. That emotional attachment probably wouldn’t be satisfied by a machine,” she says.
However, Ravichandran notes that although her followers don’t engage with fashion in that way, there’s clearly a large audience that does. “Those are the same people using Pinterest and TikTok to search outfit ideas. It’s all part of the same archetype — and I think that’s what we’re seeing evolve here,” she explains, adding that it could exacerbate the already problematic speed and churn cycle.
In a culture increasingly defined by convenience and acceleration, the appeal of generative AI makes sense — it’s fast, frictionless and always available. But what gets lost in that frictionless process is often the most meaningful part: the dig, the detour, the slow cultivation of taste. Whether the majority of the population cares about that is, however, debatable.
Comments, questions or feedback? Email us at feedback@voguebusiness.com.
I asked ChatGPT to find me a responsibly made white T-shirt. Here’s what happened
