Adobe Firefly Design Intelligence — build brand value through consistent creative expression across every global touchpoint.
10-28-2025
Anyone who manages an enterprise brand will tell you that a brand is greater than the sum of its visual parts. The way a brand shows up in the market is the direct result of years of thoughtful application of explicit rules, interwoven with subtle design nuance and best practices. Consistently delivering that level of nuanced execution across touchpoints, interactions, and geographies is what distinguishes highly ownable and distinctive brands — ones that make their stakeholders feel something — from those that have just checked the box of developing a logo, typography standards, and a color palette.
It has become increasingly challenging to scale experiences well across multiple brands, an even greater number of campaigns, and multiple different geographies — each with their own needs and execution partners. While AI has helped countless organizations accelerate and improve the first mile of creative ideation, few, if any, have met the exacting needs of enterprises when it comes to the last mile of production-ready content.
With this in mind, Adobe embarked on a focused journey with The Coca-Cola Company, creating what we called Project Fizzion. Project Fizzion aimed to solve these exact needs, serving the high design and brand identity standards of the organization while supporting scale across creative teams, partners, and agencies globally. The success of Project Fizzion set the stage for Adobe to bring an all-new solution to our enterprise customers — Firefly Design Intelligence.
A new design approach for an AI-first world.
Firefly Design Intelligence is a human-led solution to solve three key challenges.
- How do we maintain brand and design consistency as content scales?
- How can we effectively help global teams navigate multifaceted guidelines?
- How can AI help designers be more creative and innovative, not less?
The way it works is simple — creative directors within a brand, who are often but not always senior design leaders sitting at headquarters or agencies of record, can train StyleIDs, machine-readable identities which capture and convey the overall design aesthetic as well as specific design rules for a given brand or campaign. They do so directly in Adobe applications such as Illustrator, within an intuitive trainer panel UI — meaning minimal task switching or learning how to use a new piece of software from scratch.
In-house creative and agency teams then access these StyleIDs in the same tools, leveraging the intelligence and direction contained within them to create on-brand, contextually relevant designs to serve their market, channel, or campaign. They can generate new designs to reflect different themes, with the system selecting the right preapproved assets within approved layout constraints and generating contextually relevant copy.
Teams can then create a near-infinite set of new asset sizes, with the system automatically reflowing aesthetically pleasing, design-consistent layouts to match.
Four unique benefits.
Because Firefly Design Intelligence exists in the applications creative teams already know and trust, many of its AI-powered capabilities are particularly intuitive and easy to master. Firefly Design Intelligence allows teams to:
- Train for fast, on-brand execution. Train, test, and share brand compliant design systems in minutes, without leaving the tools creative and design professionals use every day. The training is easy, with no expertise in AI needed.
- Rapidly generate flexible design variants. Make traditionally static design objects intelligent with fully editable size variations, using brand-approved assets and prompts. This harkens back to the question Coca-Cola asked at the outset of Project Fizzion — “What if our logo was intelligent enough to understand how it should behave within a design?”
- Auto-select assets that conform to designs. Automatically select assets as new themes within the campaign are ideated, all from a curated set of options and with the ability to upload new assets as needed. For example, changing the theme from Pizza Party to Taco Tuesday would trigger a selection of the most relevant assets. This translates to time savings for creative teams, while enabling a greater breadth of output.
- Generate context-aware copy. Create and regenerate on-brand copy to match shifts in campaign themes while maintaining contextual awareness. In the example above, the shift from Pizza Party to Taco Tuesday triggers new copy generation relevant to the latter, which is then paired with the newly selected assets.
Anatomy of a StyleID.
StyleIDs are elaborate and comprehensive without being complex. They contain four categories of elements.
- Global brand-level elements. The most fundamental components and rules governing a brand’s visual and verbal identity system, including brand colors, logos, editorial voice, and personality.
- Campaign-level elements. Like the above, these define the visual and verbal identity system for a campaign. They may include specific formulas for campaigns, such as “Always stack the text over three rows. Follow composition of short snappy callout text.”
- Design-level elements. These inform the contextual display logic to reflow text and assets for all sizes. This is where Firefly Design Intelligence really shines — it is able to capture highly nuanced design rules such as how individual elements should lock up together (e.g., as logo plus editorial), contextual associations (e.g., types of images and copy that belong together), and visual focus definitions (e.g., locking in on humans as the subject matter in an image).
- Reference elements. These are hero templates that serve as layout benchmarks, guiding and informing the reflow capabilities of Firefly Design Intelligence as designers create new layouts for whichever aspect ratio they may need.
We’re excited to bring Firefly Design Intelligence to market as an AI-powered, human-centered design system, helping you and your global teams scale while staying on brand and compliant.
Aaron Mitchell Finegold is the head of product marketing for Adobe Firefly Enterprise, a suite of generative AI offerings that include Firefly Services, Firefly Custom Models, Firefly Design Intelligence and Firefly Creative Production. Previously, Finegold was the Chief Marketing Officer at Kingsley Gate, a professional services multinational, where he developed large and small language models. Earlier, he served as a business operations leader at LinkedIn, an associate partner at McKinsey & Company, and a strategy consultant at Ogilvy & Mather Worldwide. He has been published in Harvard Business Review, McKinsey Quarterly, and Thrive Global. He holds an MBA from INSEAD.
Recommended for you
https://business.adobe.com/fragments/resources/cards/thank-you-collections/firefly
