How real design studios
actually use AI.

Chuck Studios, an Amsterdam-based creative production company specialising in food and drink imagery, treats AI as a helpful-but-flawed tool for rapid iteration and generating variations at scale rather than a replacement for photography. It runs a hybrid workflow: shooting real ingredients in-camera first (to respect physics like viscosity and elasticity that AI can't grasp), then generating AI-led variations on top. Rather than chase every platform, it built its own front-end tool, Chuck.OS, that sits on top of the best AI tools to let staff generate on-brand images easily. The founder is skeptical AI can render food's sensory 'deliciousness' and argues that how an image is made is ultimately irrelevant to brands as long as it works.

“Whether that data is captured through a lens or generated by AI ultimately makes no fundamental difference. What matters is the effect the images have on helping brands grow. The 'how' is irrelevant.”

— Olaf van Gerwen, founder and global creative director, Chuck Studios

FutureBrand, a global strategy and design agency, operates under a formal AI policy and 'be transparent, be ethical, be human-first' principles, using AI only where it adds value across four stages: ideation, generation, validation and governance. Strategists surface early signals and test ideas from new angles; designers explore directions and prototype on-shelf designs quickly - but AI supports rather than leads. It has developed proprietary tools including Validator (stress-tests early ideas against synthetic audiences) and OnBrand (near-real-time brand-consistency feedback), plus multi-agent orchestration. Its stance is enthusiastic but firmly human-first, insisting taste and strategic vision (its bold Merck rebrand, which AI would have flagged as too risky) remain irreplaceable.

“The most powerful AI tools are the ones that feel like extensions of your process, not interruptions to it. That's when AI stops being just another machine in the room and starts to feel like a true collaborator.”

— Stephen McGilvray, executive creative director, FutureBrand

Your Majesty, an Amsterdam/London/New York brand, design and emerging-tech studio (part of Infinum), gives staff 20% of their time for AI experimentation and runs a dedicated AI unit governing tools and data (no client data used for training). It uses AI heavily at the front of the creative process for rapid exploration, plus admin work, and teaches two regular masterclasses. Named tools include NotebookLM for stitching strategy material, AI-native coding tools Replit and Lovable for prototyping, and its own Brand Delta Analysis tool that benchmarks a brand's perception across major AI models to surface hallucinations. Both leaders are excited but stress AI struggles with cultural nuance, so human curation and art direction remain essential to avoid a 'sea of sameness.'

“AI struggles with lateral thinking. It becomes effective when you give it two strong inputs and ask it to merge them. I compare it to DJ-ing. You choose the tracks and AI does the mixing. Interesting things sometimes happen in that crossfade.”

— Georgios Athanassiadis, managing director, Your Majesty

JKR, a London/New York branding agency, runs a client-driven 'AI playbook' of approved tools and policies often embedded in client contracts, backed by in-house legal and IT teams. It treats concept visualisation and mood-films via Midjourney and ChatGPT as 'table stakes,' but argues the real value is in-market: its Coca-Cola 'Adobe Fizzion' project produced custom AI style guides to keep the design system consistent across global partners. It is rolling out training co-created with Adobe around Firefly and piloting ChatGPT for Enterprise. The CEO is excited but champions the rising value of 'real' craft, arguing clients do care how work is made and human hands remain among the most valuable tools.

“The first time my Mum told me over Sunday lunch that she had 'ChatGPT'd it' was when I realised AI is now having the same impact on brands as search and social have had over the last two decades.”

— Jonny Spindler, UK CEO, JKR

Mutabor, a Hamburg-based branding agency (clients include Porsche, Montblanc and the Bundesliga), embraces AI as 'another material to work with' and stays deliberately technology-agnostic to avoid model lock-in. AI runs through research, concepting, visual exploration, content production and workflow automation, and it built its own on-brand image tool, Mutabor.AI, on clean, licensed data offering enterprise clients copyright indemnification and EU AI Act compliance. Its stack includes node-based ComfyUI and Weavy, and Langdock (unifying major LLMs under EU data processing). The stance is excited but 'eyes wide open': AI handles speed and iteration, humans remain responsible for context, emotion and killing work that doesn't feel true to the brand.

“We tell our teams, 'Easy tasks no longer exist. Hard tasks are the new easy tasks. Impossible tasks are the new hard tasks.'”

— Mutabor (attributed to the studio)

Among Equals, a London strategy and branding studio, treats AI strictly as a 'thinking partner, not a creative' - using it for quick research, generating visuals and moodboards for pitches, sense-checking tone, and interrogating its own thinking like a coach. The team is entirely self-taught (via internal 'Lunch & Learn' sessions) and also leans on AI for back-office work: HR policy, finance modelling, tightening contracts. For deep research they use Claude, image generation across Nano Banana, Endless Tools and Visual Electric, and interactive prototyping in Figma Make. Founder Emily Jeffrey-Barrett is 'cautiously curious,' insisting AI can level the playing field for small agencies but can't imagine, decide, or replace human judgement.

“It's a valuable thinking partner, a powerful way to bring ideas to life. But relying on AI as you would a human is a dangerous game. Outsourcing your brain to an app is like plating up Deliveroo every night and pretending you cooked.”

— Emily Jeffrey-Barrett, founder, Among Equals

View Source, a New York/Brooklyn design and technology studio, set a policy that AI only enters after humans set the creative direction - used for conceptual thinking, moodboards, research and 'agentic efficiencies.' Creative director Andrew Rutledge personally trained daily for a year, from Stable Diffusion to ComfyUI, and the studio trains its own custom AI models for tighter control. Concrete work includes an 'otherworldly' rebrand for blockchain platform Moonbeam (Stable Diffusion creatures and landscapes turned into controllable motion assets) and rapid prototyping of site animations via ComfyUI. They stress owning your input data for creative integrity.

“For us, AI only comes in after people set the direction. We shape the vision, craft the imagery, write the strategy, and then use AI to create agentic efficiencies. That's how we keep control, maintain a high degree of taste, and stay differentiated.”

— Andrew Rutledge, creative director and partner, View Source

Studio Dumbar, a Dutch branding, design and motion agency and part of DEPT, is excited about AI as an extension of its long history of building its own tools through creative coding. It uses AI for brainstorming, generating images/video for early inspiration, coding experiments and upscaling, plus back-office tasks. Use is governed through a DEPT Digital Ethics Advisory Board, global AI guardrails and client AI policies - notably ~80% of clients forbid AI in the final deliverable over legal concerns. They favour tools that keep the designer 'steering,' such as node-based Flora, and are experimenting with training their own models in Automatic1111 and ComfyUI.

“It's about putting AI into our workflow, not turning the workflow over to AI.”

— Studio Dumbar (answering collectively)

Base Design, a global creative network (New York, Brussels, Geneva, Melbourne), is unreservedly excited about AI, framing it as design's next chapter after the printing press, Photoshop and Figma. It sees AI's real value in exploration - pushing thinking, testing ideas in multiple directions, simulating scenarios and rapid prototyping ('vibe coding') - while guarding against homogenisation with a living AI policy built on transparency and client-IP protection. Its Base Digital engineering group are the heaviest users, and the studio connected shared drives and Slack to a private GPT for knowledge management. Partner Mirek Nisenbaum lists a broad toolset: NotebookLM, Flora, Midjourney, Firefly, Runway, Flux Kontext, Krea, Cursor, Topaz and more.

“Our job is to bring discernment, taste, and critical thinking to the table, ensuring AI pushes us beyond the common denominator rather than dragging us down to it.”

— Mirek Nisenbaum, partner and digital director, BaseNYC

Studio Gruhl, a Berlin strategy and design agency, has worked with AI since 2019 (including training its own early models) but is notably skeptical, viewing it as 'still statistics, not true intelligence' and warning of an economic bubble. It uses AI as a support to accelerate workflow - sketching ideas, quick iterations, setting a visual mood, upscaling, and shortening render times in 3D - while noting it breaks down on detail, consistency and anything outside common visual cliches. They are ethically deliberate: writing prompts from the ground up, avoiding artists' names and unowned reference images. Named tools are Krea (upscaling), Midjourney (visuals and short motion tests) and Black Forest (their own datasets).

“AI has made huge leaps since then. But the reality is that it's still statistics, not true intelligence. So AI is a support, but not a tool by itself.”

— Malte Gruhl, founder, Studio Gruhl

Buck takes a deliberately software-agnostic, 'middle-out' approach to AI, treating it as a sandbox for exploration rather than a shortcut to final output. In production it uses generative tools during ideation and prototyping (grounded in human sketches and line tests), speeds up repetitive 3D tasks, and recommends Krea for playful testing and ComfyUI for structured, production-quality workflows; adoption began organically when a Midjourney Experiments channel exploded on its Slack. Beyond image-making, Buck is building internal 'assistants' for talent resourcing, client-brief analysis and audience modelling, envisioning brand systems evolving into a 'Brand Design Brain.' The stance is enthusiastic but wary: humans stay in the driver's seat.

“We very much believe Humans, with a capital H, should stay in the driver's seat. Clients want and need to collaborate with people they can build trust with. That relationship can't be automated.”

— Vincent Lammers, Executive Creative Director and Partner, Buck

&Walsh uses AI mainly in the early stages of projects for moodboarding, rapid ideation and generating unexpected references to push past creative blocks, while insisting it can't handle nuance, strategy or cultural specificity. The team treats experimentation as a 'collaborative playground,' sharing prompt techniques rather than following a formal policy, and also leans on AI for admin like organising research and summarising notes. Recently it has focused on video-generation models, notably Google Flow, to gather references and quickly build TVC animatics. The guiding principle: AI supports original thinking rather than substituting for it, with an eye on ethics and authorship.

“We're in the process of shaping a more formal policy, but the guiding principle is this - we only use AI when it supports our original thinking, not as a substitute for it.”

— Jessica Walsh, Creative Director, &Walsh

OHMY encourages everyone in the studio to use AI freely, governed by a simple honesty rule: use it as much as you want as long as you'd admit it. It applies AI across deep research, fleshing out and reviewing concepts, spotting gaps in strategy, coding, day-to-day formatting, and built internal tools for costing and estimating. Co-founder Joe Burke points to the ChatGPT release that could suddenly render images and typography as his turning point, and relies on the dictation tool Talktastic to feed richer context into prompts. The studio is careful legally, paying for subscriptions so client content isn't used for training and updating contracts to remove generative-AI liability.

“The simple rule is - use it as much as you want, as long as you're happy to admit it. If you'd be embarrassed to say you used AI on something, you probably shouldn't.”

— Joe Burke, Co-founder and Creative Director, OHMY

COLLINS frames AI as just another tool in a long lineage of disruptive technology (co-founder Brian Collins draws a parallel to the 1985 Macintosh and Emigre fonts), arguing it will make mediocre work more mediocre but let skilled designers make extraordinary work. The studio has roots in generative, coded design systems going back over a decade (algorithmic visuals for Spotify), and today uses AI for agentic systems, research, visual examination, content exploration, workflow acceleration, project management and data analysis. Collins pointedly refuses to recommend any specific tool, insisting the tool is irrelevant next to human judgement. The stance is one of relief rather than excitement or nervousness.

“Let's be clear: AI is a tool. It is not a creative director. We use it where it makes sense - agentic systems, research, visual examination, content exploration, workflow acceleration.”

— Brian Collins, Co-founder, COLLINS

Superside, a Berlin-based design, branding and creative service, has AI integrated across its full creative process: ideation and research, moodboarding and early visual exploration, image manipulation ('Photoshop on steroids'), and custom AI-driven workflows that auto-populate design templates and plug into Figma. It runs on a formal set of AI principles (opt-in/opt-out flexibility, human-guided creativity, data security) and trained its entire 300+ creative team over six months. It discusses AI up-front with clients, who can opt in or out per project or account. Its stance is one of clear excitement, viewing AI as the next chapter of progress while insisting taste and creative direction stay human.

“The one thing AI isn't great at is taste and judgment. It's a powerful assistant, but creative direction still starts and ends with humans.”

— Kae Neskovic, Executive Creative Director, Superside

Hey Studio, a Barcelona-based creative agency, uses AI mainly to prototype ideas it already has and occasionally for illustration exploration (lights, textures, finishes) - strictly for prototypes, never final deliverables. It has a cautious policy written into its contracts refusing briefs that arrive with AI-generated art direction, and favours self-directed learning via a weekly 'Hey Lab' every Thursday where the team experiments on imaginary projects. Founder Veronica Fuerte also uses AI as a business 'coach' and sounding board. Her stance is ambivalent but adaptive - AI as both exhilarating and disruptive.

“Right now I feel like I'm somewhere between paradise and hell. On one side, it's a beautiful moment because AI pushes me to be better, faster, and even a stronger designer.”

— Veronica Fuerte, Founder and Creative Director, Hey Studio

Taxi Studio, a Bristol-based brand and design agency, uses AI mainly to visualise thinking at speed - accelerating the early visualisation stage so the team can spend more time on ideation and problem-solving. Inspired by Asimov's Three Laws, it wrote its own 'Five Principles of AI' in March 2024 for guardrails and client transparency, and partnered with the Pratt Institute for an intensive course on prompt writing, custom GPTs, Midjourney and Firefly (a 'Creative AI in Design' accreditation). Adoption traces to 2023 when Midjourney went public. The stance is excited but nervous - AI will take over production tasks, but human empathy and gut instinct stay irreplaceable.

“For now, one thing AI can't replicate is empathy, understanding human emotion, or having that gut instinct that tells you, 'Something about this doesn't feel right.'”

— Stu Tallis, AI Lead, Taxi Studio

MOTTO, a New York-based strategy and design firm specialising in tech and innovation (with people in London, Berlin and Dallas), uses AI as an early-stage 'co-pilot' to stretch imagination across moodboards, naming, storytelling prompts, visual studies and brand-voice development, and behind the scenes for documentation and content repurposing. It built a proprietary Motto Perception Map in 2024 that blends its methodology with AI to evaluate brand performance across seven dimensions, and maintains a formal legal AI policy plus an L&D library on prompt design, image generation and AI ethics. Its stance is thoughtfully pro-innovation - AI as a sparring partner that sharpens thinking but never replaces intuition or taste.

“We use AI as a teammate and sparring partner. It sharpens our thinking, but it doesn't do the thinking for us.”

— Sunny Bonnell & Ashleigh Hansberger, co-founders, MOTTO