A new era of technology is upon us. Generative artificial intelligence, also known as GenAI, has been making waves in the tech world. Just as we thought we’d seen it all with AI and its ever-expanding tools, it has now emerged as the launchpad of what feels like an entirely new playing field. Particularly in recent years, AI has seen rapid growth, with its use extending across a diverse array of fields. One massive sector that has felt the impact of this technological revolution? The fashion industry.
From enhancing consumer experiences to predicting trends, AI’s role in the space is constantly evolving. Its latest venture dives into the creative process with tools that can now be harnessed for design and production operations. Not to mention the advantages it offers in terms of efficiency and cost: projects that once required a large team can now be executed with fewer resources, saving both manpower and money. Another notable factor is its ability to ‘make the impossible possible’, pushing creative boundaries by generating images and concepts that were previously unimaginable.
Enter Maison Meta, the world’s first GenAI agency. Founded by Cyril Foiret about 2 years ago, the venture encompasses everything from creative consulting and GenAI campaigns to research and development, as well as conducting training and workshops. The French founder comes from a creative background where he began his career as a wardrobe stylist before creating Trendland, an online platform covering trends in fashion, art, photography, design and culture.
Maison Meta launched AI Fashion Week, a global competition uniting a community of fashion enthusiasts and tech innovators. This groundbreaking initiative was met with immense success, leading to a second season in partnership with Revolve. It offered participants a unique platform to showcase their talents and explore the possibilities of merging fashion with cutting edge technology.
As we move further into this new era, the impact of GenAI on fashion is only set to grow. With trailblazers like Maison Meta at the helm, the convergence of creativity and technology will keep pushing the boundaries of innovation, setting the stage for new advancements. The future isn’t just about fashion evolution. It’s also about how technology will drive the next wave of digital transformation, redefine industries and reshape the way we think about art.
Here, Foiret and his business-development colleague, Nima Abassi, share with Vogue Singapore about merging AI with fashion, their most memorable projects and their vision for the future of technology in the fashion industry.
Let’s dive into your journey in the digital space. How did you start and what inspired you to use AI over other digital tools?
Foiret: I was eager to use AI as a new technology to show the fashion and beauty industries what’s possible. We embraced the tools as soon as they were available, focusing on research and development and pushing them to their limits. Having collaborated with major brands and renowned photographers, we saw AI’s potential for brands, artists and clients, particularly in fashion.
Abassi: Cyril and I met during the Web 2.0 era, when social media was on the rise. Cyril’s background started in wardrobe styling in Paris, and he later founded a popular online magazine focused on trendspotting. I came from a different path—first as a lawyer, then as a management consultant—viewing new technology from a business perspective, while Cyril looked at it creatively. Combining our approaches made sense with AI as it’s a powerful tool for both creativity and business. The adoption has been rapid; in the past year and a half, it’s evolved considerably, with fashion brands increasingly interested in using it to meet the growing demand for quality content across various platforms.
Could you walk us through your creative and design process when working on AI projects?
Foiret: We have different areas at Maison Meta: a studio for GenAI creative work, consulting focused on AI strategy for brands, and workshops and training for clients worldwide. We’re also now teaching AI in academic programmes, like the upcoming AI course for Istituto Marangoni’s MBA programme, and we’ve recently been confirmed to teach at Savannah College of Art and Design. For creative projects, we start with a brief where clients provide direction on style, mood and ambience. Our role is to translate their vision using AI tools, sometimes with creative freedom and other times with precise requirements, such as incorporating real clothing into the images.
After creating the initial AI-generated images, we refine them using Photoshop, much like traditional editorials, to adjust details, correct mistakes and fine-tune the final look.
How has AI changed your approach to experimentation and pushing creative boundaries?
Foiret: We often discuss the concept of the ‘impossible shoot’—creating visuals that would be logistically or financially challenging in real life. AI allows for exaggerated and imaginative scenes that push creative limits. For example, in a campaign with Pangaia, we depicted over-the-top visuals like a lady on a scooter surrounded by a mountain of oranges. AI opens up possibilities for experimenting beyond traditional constraints, whether it’s for advertising or inspiration boards.
AI also educates clients on what’s feasible and what isn’t. There’s a misconception that generating stunning visuals is as simple as typing a prompt, but there’s significant work involved in layering and refining images to meet client expectations.
Let’s talk about AI Fashion Week. What inspired this initiative and were there any standout moments?
Foiret: AI Fashion Week started because we wanted to be the first to showcase AI’s capabilities in fashion. We brought together the AI community through Instagram and Discord, encouraging participants to design collections with consistency and creativity. It was an exciting first step and we’re pleased with the results.
The competition revealed incredible talent from around the world, with over 450 submissions in the first season alone. It was inspiring to see how many people embraced this new tool to showcase their creativity. The samples will be available very soon and are expected to go on sale on Revolve’s website.
How would you say AI Fashion Week is different from a traditional Fashion Week?
Foiret: Well, for now, it’s more of a competition with many designers. The idea is to give AI artists a chance to potentially become designers and maybe launch a fashion brand at the same time. Revolve is also helping them by training them how the business works and guiding them into the fashion world. It’s different because we don’t have a full runway, which we hope to introduce in the future. Also, some of the artists aren’t traditional fashion designers; they come from different backgrounds, have a fashion sensibility and can use the tools creatively.
Abassi: As you know, producing a collection and staging a show in the real world involves massive costs, especially in places like New York and Paris. AI removes that financial barrier, enabling more people to showcase their creativity without traditional obstacles.
What are some of your favourite or most notable projects?
Foiret: The first one with Moncler was pivotal; it got us noticed and brought in some press. Moncler was interested in seeing how AI would interpret their collection, not replicate exact pieces. Also, collaborations with Pharrell [Williams], Adidas Originals, Alicia Keys and others each had a unique vibe, which we created using tools available at the time, along with retouching. We’ve also done some fascinating but unreleased tests. For instance, we worked on creating big models based on Norma Kamali’s dataset, allowing her to use the tool to design new collections.
Regarding Norma Kamali, we developed an AI model based on her 40-year archive, training it to understand her design DNA. Now, she uses this tool to create consistent new collections, and the AI understands her brand’s look, whether it’s swimwear, dresses or jumpsuits.
“AI allows for exaggerated and imaginative scenes that push creative limits.”
What are some challenges you’ve faced along the way as Maison Meta has grown?
Abassi: One big challenge is the general lack of understanding among clients about what AI can and can’t do. However, this also presents an opportunity for us since few people in fashion truly grasp the intersection of creativity, technology and business.
Foiret: On the technical side, we initially struggled with computing power; AI work requires high-performance PCs, not typical Macs. Output quality was another issue; early images weren’t defined enough for large formats such as billboards or magazines. Now, the challenge lies in creating videos. Brands want high-quality video content comparable to professional videographers, which AI isn’t fully ready to handle yet, but it’s improving.
What are the advantages of using AI to create images?
Foiret: It’s primarily about efficiency and cost, which is why many brands are adopting AI. It also offers more control, allowing brands to bring these processes in-house, especially for faster marketing materials or lookbooks. While the technology isn’t fully developed for e-commerce yet, we’re seeing significant demand from brands wanting to create e-commerce imagery with AI to quickly update their sites and branding. Ultimately, it’s about saving time, cutting costs and improving efficiency.
Abassi: You can use AI to make content that’s relevant to your audience in different locations, greatly enhancing efficiency. It also brings a creative angle, enabling you to imagine things you might not usually consider or create scenarios like the ‘impossible shoot’ we often discuss. Essentially, it allows you to produce content that would otherwise be impossible or extremely expensive to create.
How do you think the use of AI will evolve in the near future?
Foiret: We’re currently just at the beginning. Last year, most fashion brands were still exploring and trying to understand AI. Now, they’re looking to integrate it into their workflows. This is just the start. It’s going to be everywhere, and as the quality and ease of use improve, it will significantly transform the industry.
Abassi: I think the ease of use and user interface experience are areas that will see significant improvements, which will drive wider adoption. We’re currently developing a generative AI app—a tool that codifies our workflow and the tools we use, something clients frequently enquire about. We plan to launch this within the next six to nine months. This will make it easier for organisations and individuals to use the tool as their creative outlet, greatly supporting broader adoption.
What’s next for Maison Meta?
Foiret: We’re preparing to launch our new tool, Seeed.AI. This app is designed with built-in workflows and specific use cases tailored to the fashion industry, which has one of the highest demands for high-quality content. By bringing our technology in-house, fashion brands can replicate what we do on a much larger scale. Since we’re a small team and not many are familiar with pushing the boundaries of GenAI as we do, our growth and project capacity are limited. With this tool, however, we envision that hundreds, if not thousands, of fashion brands will be able to create exceptional content in-house at a fraction of the time and cost. We don’t expect traditional content creation such as photoshoots to disappear, but Seeed.AI will serve as an invaluable addition, allowing brands to work more efficiently and creatively.
Additionally, we’re developing what aims to be the most versatile and comprehensive AI fashion model. We’ve been training this model on licensed datasets and feeding it detailed knowledge of fashion, such as seams, pockets, fabrics and more. This model will serve as a reference for the industry, generating clothing details that align with the design vision. This will be integrated into our upcoming app, Seeed.AI. We’ve already started the groundwork, but it’s a significant effort to train and prepare this model fully.
The October anniversary ‘Community’ issue of Vogue Singapore is available online and in-stores now.
