There’s a lot of crappy clothing out there, and the proliferation of fast fashion to meet the demand of even faster-moving trends makes it tempting to fire up the old credit card and fill your cart with flashes in the pan made of plastic byproducts. Dressing for trends can be a losing game, especially with limited closet space and budget. So, I’m turning to artificial intelligence for help in building the only wardrobe I’ll ever need.
A quick Google search reveals a rich and competitive landscape filled with AI tools meant to guide budding fashionistas through the scrum of today’s offerings.
Style DNA, for example, promises to “assist with your everyday style dilemmas” through facial and item recognition designed to help you dress for your body type and different occasions. Stylee uses AI to generate specific looks and complimentary color palettes, and Aiuta offers direct purchasing for its recommendations.
These sound great if you’re not sure where your personal style lands and need from-scratch advice on how to dress, but I already have a good idea of what I’m looking for. Capturing the nuance of my split-personality style is a challenge I suspect only one tool can tackle: ChatGPT.
The right tool for the job
ChatGPT, launched in 2022 by OpenAI, was recently updated with improved writing, math and logical reasoning. Its training data has been updated, meaning you can get information as recent as January 2022 for GPT 3.5 and April 2023 for GPT-4 Turbo. Not quite real-time search but getting closer. (You can also check out our reviews of Gemini, Claude and Microsoft Copilot on CNET’s AI Atlas resource page.)
This freemium AI tool operates using conversational — sometimes complicated — search prompts to return results from its training data. Kind of like a chatbot that pulls its information from limited internet access instead of a preprogrammed set of responses.
You can either make a limited amount of requests each day for free, or pay anywhere from $20 to $25 a month to upgrade to ChatGPT Turbo and increase the number of requests. It depends on how chatty you want to get.
For the advice I want from ChatGPT, I have pretty specific criteria:
No plastic clothes. The invention of the polyester blend has bloomed into the bane of conservationists shouting at consumers about the use of plastics in fast fashion. I’m looking to eliminate plastics from my wardrobe — mostly due to a traumatic sweat-related incident during an important pitch meeting in which my nonbreathable polyester top thoroughly betrayed me — so I want items with at least 90% natural fibers. Classic looks. Since we’re shooting for a wardrobe built to last through all the low-rise, mom-jean, wide-leg, skinny-pant, bold-print whiplash trends, I want to stick with items that will wear just as well in my elder millennial era as they will when I’m an actual elder yelling at AI-generated teens to get off my cyberlawn. Modest price point. While I’m fortunate to have enough money for this experiment, I’m also not so fortunate to have been born extraordinarily wealthy or have a long-lost relative leave me millions in their will, so I’ll be giving ChatGPT a budget. Versatile players. My work takes me from the classroom to content shoots, to around the country giving talks and holding workshops at conferences, to pitching marketing ideas in the boardroom, but on the weekends I lead a double life as a grade-A slob. These clothes will need to be comfortable enough to bum around in and not so precious that I can’t spill a little wine on them without freaking out.
Having so many specifics makes using one of the AI stylist tools pointless: The in-app shopping features don’t allow me to specify what material to avoid or incorporate. So it’s time to find out if ChatGPT has fashion sense.
Does ChatGPT understand style?
I started my AI-assisted fashion journey by telling ChatGPT what I was looking for with those specifics, as well as some of the items I already have and notes on color palette.
The tool returned with a smattering of suggestions and price points, as well as some picks on where to purchase the items.
Screenshot by Rachel Kane/CNET Screenshot by Rachel Kane/CNET
This was a fine start, but plain white tees and versatile cardigans are pretty generic. I decided to try something a little controversial and asked ChatGPT to role play as celebrity stylist Law Roach, whose partnership with Zendaya has resulted in some of the most iconic looks of the last decade. Plus, I just saw Challengers, and Zendaya’s chic country club-adjacent outfits have been haunting my dreams.
Match points
Now armed with the outlook of Roach, ChatGPT repeated its suggestions but added more detail, offering styling tips like high-waisted trousers and slightly oversized fits for the cashmere sweaters.
It’s getting closer to giving me what I want, but there is still the matter of accessories and how the items I already own can work together with these new items. I asked for suggestions on how to pair what I have in my closet with what I’m buying.
This is where ChatGPT started to shine.
The results of my updated queries not only produced useful and achievable looks to build with my current wardrobe, but also gave me working links for items in the price range that met my criteria. Pretty impressive for a tool that had previously been unreliable most times I tried to use it.
Screenshot by Rachel Kane/CNET Screenshot by Rachel Kane/CNET
The verdict on AI as a stylist
If you’ve got specific criteria for any shopping endeavor, I’d recommend trying ChatGPT. The tool cut out tons of time I would have dedicated to scouring the web for things that might work from a style standpoint, and it did all the heavy lifting in making sure the items were mostly made from natural fibers.
The benefit of having the tool act as a known stylist with a particular point of view was extra helpful in determining how to make certain pieces fit together in an appealing way.
As for actually looking like Zendaya in the outfits, well, that’s a challenge AI can’t help this human overcome. Not yet, anyway.
Editors’ note: CNET used an AI engine to help create several dozen stories, which are labeled accordingly. The note you’re reading is attached to articles that deal substantively with the topic of AI but are created entirely by our expert editors and writers. For more, see our AI policy.