Methodology
The criteria we apply, what we look for, and the principles that govern which agencies appear on this site.
How we evaluate agencies
Every agency on this site has been assessed against five criteria. No numeric scores are assigned — the criteria determine whether an agency meets the threshold for inclusion and how it is positioned within the directory. The central challenge is distinguishing genuine AI capability from positioning. These criteria are designed to separate documented practice from marketing language.
Whether an agency's AI use is verifiable, not self-reported. We look for named tools at named stages, case studies that describe AI's role specifically, and client feedback referencing AI-assisted delivery. Agencies that describe AI in general terms — "we use AI throughout our process" — do not meet this criterion. We distinguish commercially available tools from proprietary AI infrastructure; both can qualify, and the distinction is noted in profiles.
Strong AI claims mean nothing without strong underlying work. We evaluate through live deployed projects, not case study PDFs — looking for output still in active use by credited clients, coherent across the surfaces those clients operate on, and durable enough to hold up as companies scale.
AI in design covers a wide range of actual capabilities — production acceleration, research and insight, brand system generation, and AI product design are fundamentally different practices. We assess whether each agency's AI capability is coherently matched to the clients and projects it claims to serve, and note explicitly where an agency is strong in one mode and absent in another.
AI workflows introduce specific governance questions: how AI output is reviewed before delivery, how brand consistency is maintained across AI-generated assets, and how data handling and IP ownership are managed. We evaluate whether agencies have documented answers, and weight client feedback from identity-verified platforms alongside self-published process documentation.
Third-party signals: verified client reviews referencing AI-assisted delivery, industry recognition for AI-integrated work, and editorial coverage addressing the agency's AI practice specifically. Verified reviews from named clients on Clutch carry more weight than website testimonials. AI-specific validation is still developing as a category — agencies with strong general design validation but limited AI-specific recognition are not excluded.
What we don't evaluate on
Listing Midjourney, ChatGPT, Figma AI, and Runway on a services page is not evidence of AI capability. What matters is the process built around the tools, the guardrails, and the institutional knowledge that determines whether the tools produce reliable, on-brand output.
Some of the most rigorously AI-integrated studios in this directory are small. Team size is noted in profiles as context, not as a selection input.
This is a global directory. AI-forward design practice is geographically distributed — Portland, Berlin, Amsterdam, Victoria BC, and San Francisco all appear in these profiles. Geography is noted and is not a factor in selection or positioning.
Being early to AI adoption is not the same as having built meaningful capability. We assess the depth and documentation of current AI practice, not when the agency first mentioned AI on its website.
An agency that describes its AI practice extensively in marketing materials but has no verifiable client feedback, case study detail, or third-party documentation referencing that practice does not meet the AI workflow documentation criterion.
Selection threshold
Inclusion requires a minimum level of portfolio quality, at least one verifiable example of documented AI workflow integration, and sufficient public evidence to evaluate across all five criteria. Agencies that position themselves as AI-forward but cannot provide verifiable documentation of that practice are not listed regardless of their general design quality.
The directory is reviewed once per year. Individual profiles are updated on a rolling basis when significant new work, documented AI capability changes, or new validation signals emerge. Factual corrections can be submitted with supporting documentation.
FAQ
Agencies are identified through our own research process. If a studio is not listed, it either did not meet the inclusion threshold, lacked sufficient documentation of its AI practice, or has not yet been reviewed.
Because this directory exists specifically to cover agencies using AI in their design process. An agency with an exceptional portfolio but no documented AI practice belongs in a general design directory, not this one. Portfolio quality matters — it's the second criterion — but it's secondary to the core question of whether the AI positioning is substantive.
Through evidence that exists independently of the agency's own claims: case studies that describe AI's role with specifics (which tools, at which stage, what the output looked like), client reviews that reference AI-assisted delivery, published explanations of AI infrastructure, and work that demonstrates the kinds of output — volume, variation, consistency — that AI-assisted production actually produces. Agencies that can only point to their own marketing copy as evidence of AI capability don't meet the documentation criterion.
No — the five criteria apply equally regardless of whether an agency uses AI for production, research, brand system generation, or AI product design. The type of AI use is noted in profiles and informs positioning, but it doesn't advantage or disadvantage an agency in the evaluation. A modest production-focused AI workflow that is well-documented and clearly delivered scores higher than an enterprise AI platform claim that lacks client-side verification.
Numeric scores imply a precision that isn't warranted, and in a category where the underlying capability is this variable, a ranked score table would suggest comparability between agencies that are actually doing fundamentally different things. A 9.1 for an agency doing AI-assisted brand production and a 9.1 for an agency doing regulated healthcare AI research are not the same number in any meaningful sense. Profile-level positioning with explicit Best for and Not a fit for guidance is more useful than a table.
The directory is reviewed once per year, but individual profiles are updated on a rolling basis when significant documented changes occur. An agency that has meaningfully upgraded its AI infrastructure, published new AI-specific case studies, or accumulated new client validation for AI-assisted work can have its profile updated between annual reviews. The standard for updating is the same as for inclusion: verifiable documentation, not marketing announcements.
Reviews on platforms that require the reviewer to verify their identity and company affiliation — primarily Clutch and G2. Reviews that specifically reference AI-assisted delivery are weighted more heavily than general satisfaction reviews for the purposes of this criterion. Testimonials on the agency's own website without corroboration are not counted.
Either they did not meet the documentation threshold for AI workflow integration, their AI positioning lacked verifiable substance, or they have not yet been reviewed. Reputation and client prestige are not substitutes for documented AI practice in this directory.
The directory
Each profile describes specifically how AI is used, who the firm is best for, and where it isn't a fit. Profiles never rank agencies by numeric score.
