Where AI Actually Fits in Graphic Design (And Where It Doesn’t Yet)
Categories
Where AI Actually Fits in Graphic Design (And Where It Doesn’t Yet)
As the head of the graphics team at Aventi Group, I get asked some version of this question constantly: “So… are you all using AI for design now?”
The honest answer is: yes, and no. It depends entirely on what we’re making and who we’re making it for.
Where AI Has Genuinely Changed Our Workflow
AI has become a real part of how we ideate. When we’re exploring concepts, for internal projects, early creative direction, or just trying to see a wall of possibilities before committing to a direction, AI tools are excellent. They let us generate a range of visual ideas quickly, which speeds up the front end of the creative process in a way that just wasn’t possible a couple years ago. Personally, I’ve found Claude to be a strong tool for this kind of graphics exploration.
AI has also become genuinely useful in quality control. One of the most surprising wins for us has been using AI to review final files and confirm that every requested change actually made it into a design. AI is remarkably good at catching small, missed edits that a tired human eye scans right past. It’s not glamorous, but it’s real, practical value.
Where We’re Still Leaning on Adobe
For our client work, large clients and mid-size clients alike, we still rely heavily on Adobe products. And that’s not nostalgia, it’s practicality. Adobe tools are faster for the kind of precise, brand-consistent work our clients need. They’re consistent, predictable, and they integrate tightly with established brand systems.
AI, by contrast, still struggles with branding. Getting an AI-generated output to actually match a client’s specific brand guidelines, their exact color values, typography rules, logo usage, tone, often turns into a time sink rather than a time-saver. For clients with strong, established brands who need precision and flexibility, that gap matters a lot right now.
Presentations Are Their Own World
As a product marketing agency, we work very closely with product marketing teams on decks, and that’s an area where the graphic work matters enormously. A great deck has to tell a story, and it has to tell it in a way that’s simple and immediately understandable to the audience. Our graphic artists work hand in hand with PMMs to make sure that story comes through correctly.
Anyone who has built a keynote or an executive presentation knows how much they change along the way, and how many people are involved in shaping them. That kind of iterative, collaborative, story-driven process is not something AI can replace right now. It takes a real understanding of the narrative, the audience, and the moment, and our graphic artists are exceptional at translating all of that into something visual that an audience can actually grasp and connect with.
The Photography Question
I get asked about AI photography more than almost anything else. As tempting as it is to lean into it, we’re still buying stock and commissioned photography for client content. AI-generated imagery is improving fast, but there’s also a real segment of our client base, and their audiences, that is genuinely AI-averse. Until that shifts, and until the tools offer the brand flexibility our clients require, purchased photography remains the safer, better choice.
What’s Next: Video
If there’s one area I’m watching closely, it’s video and animation. I expect this is where we’ll lean into AI more and more in the coming years. I’ve heard great things about Adobe Firefly’s video capabilities specifically, and it feels like the space where AI’s strengths, speed, iteration, generating options- could actually start to outweigh its current weaknesses around brand precision.
The Bottom Line
Net net: AI is getting better every month, and it’s already a great tool for exploring ideas and catching details humans miss. But for clients with strong, consistent brands who need precision and flexibility, traditional tools like Adobe’s suite still do the job better today. That balance is shifting, just not as fast as the headlines suggest.
