Every new AI tool promises to make creative production faster, and it's becoming harder to resist the temptation to produce more.
What once required an entire production team can now start with a few words typed into a machine.
And it's easy to see why so many businesses are getting on board.
Nearly two in three digital video ad buyers now use generative AI for creative production, compared to half in 2025, according to the IAB's 2026 Digital Video Ad Spend Report.
But here is the question brands should be asking: If AI can generate hundreds of campaign images in minutes, how many are actually worth publishing?
Behind every unusable image are hours spent generating, reviewing, and refining work that may never reach an audience.
For brands and agencies, that raises an uncomfortable possibility.
The technology promising to save time could be creating an entirely new way to waste it.
On the latest DesignRush Podcast, we spoke with cuz. co-founders Michael Bismuth and Kev Cohen about what happens when AI-generated content enters professional advertising production.
The Miami-based studio has worked with brands including DKNY, Karl Lagerfeld, David Guetta, and Jimmy Butler's BIGFACE.
And their experience challenges one of AI advertising's biggest assumptions.
"AI is cheap to generate, but it's expensive to get right," Cohen tells DesignRush.
So where does the money go?
Michael and Kev share three lessons that brands and agencies should consider before commissioning their next AI campaign:
- Why creative direction determines which AI images deserve to be published
- How hybrid production helps teams avoid expensive AI mistakes
- Where AI adds value to an existing brand campaign
Watch the full episode on YouTube or listen on Spotify.
Who Are Michael Bismuth and Kev Cohen?
Michael is Co-Founder and AI Lead at Miami-based creative studio cuz. He was a former DJ and art director, and now specializes in AI-driven visual production, with experience in music videos and creative campaigns.
Kev is Co-Founder and Creative Director at cuz., bringing over 15 years of experience in photography, videography, and design. He leads creative direction, combining traditional production expertise with AI-assisted advertising.
Choose the Idea Before Generating Images
Before discussing mood boards or visual references, Cohen wants to understand the idea behind a campaign and what the audience should feel when they see it.
"The idea is the most important because this is what will make the viewer feel something when he's going to look at the image, at the campaign, at the music video, at whatever it is," Cohen says.
Once the creative direction is established, AI can produce hundreds or thousands of variations.
Selecting the right image becomes another demanding part of production.
"Taste is what you're going to select out of a thousand options. This is what defines your taste," Cohen adds.
Bismuth shares a similar concern about brands relying too heavily on the technology itself.
"It can generate, but it can't direct. I think direction and taste (are) the most important things nowadays," Bismuth says.
For brands experimenting with AI in branding, the challenge starts before the first image is generated.
Who decides what belongs in a campaign when thousands of options are available?
Know When to Stop Using AI
One of the studio's recent campaigns involved a bicycle with a missing wheel.
What initially appeared straightforward became difficult when AI struggled to render the bicycle's physical details correctly.
Bismuth recalls the result:
"It just looked like a bunch of messed pixels."
The team eventually used a 3D model and AI-assisted lighting to complete the scene.
Their experience offers a useful warning for brands planning to rely entirely on AI production.
"Using AI for using AI is just a waste of time, most of the time," Cohen says.
Even when AI accelerates certain tasks, the production team still needs the expertise to recognize when another technique is necessary.
Bismuth acknowledges that some projects take longer when the team keeps trying to solve a problem with AI.
"But sometimes using AI will make us waste time also because we are trying so hard."
For brands assessing their video production costs, this raises an important consideration.
The budget must account for creative specialists, revisions, and the possibility of switching production techniques.
Get More Content From the Same Shoot
Cohen recalls an early BIGFACE photoshoot where he worked with the camera while Bismuth generated alternative backgrounds and props from the photographs.
The experiment helped the brand see what AI could add to an existing production.
"Michael was generating like new backgrounds as I was shooting. He was generating new props on the photos that I was just shooting," Cohen says.
A similar opportunity came with DKNY.
"They only had photos. They wanted some videos. They didn't have a videographer on the shoot."
The studio used AI to animate existing campaign photography, giving the brand video content using assets it already had.
For Bismuth, this is where brands should reconsider what they expect from AI production.
"Not always saving time or not always saving money, but just getting more outputs."
This distinction matters for agency owners pitching AI services.
What additional content could a client create from the photography, footage, and creative assets they already own?
And the commercial possibilities come with another concern.
"Nowadays, they're not afraid of AI. They're afraid that it shows. They're afraid that AI is being, that it shows," Cohen says.
Michael and Kev also discuss:
- Why brands worry about artificial-looking campaigns
- The creative skills agencies need to retain
- When physical advertising deserves another look
Their experiences raise a question for every brand investing in AI: Is the goal to produce more content, or to produce content worth using?
Don't miss the full discussion! Watch the episode on YouTube or listen on Spotify.
