Creatives are cooked because of AI ads… Or are they?

AI-generated ads have officially entered the chat.

And if you’re a marketer, it’s getting harder to ignore them. 

You’ve probably seen them by now… That Barbie AI parody. The Harry Potter “Vlogwarts.” The caveman skydiving.

AI-generated videos are flooding social feeds. Not even just as entertainment but as full-blown ads. 

That’s because they’re faster to make. Cheaper to produce. Scalable in ways traditional content creation isn’t. As of 2025, nearly 79% of eCommerce brands are using AI-generated video in some capacity.

Some are weird. Some are pretty. Some are converting. Some are all of the above.

But for all the efficiency, there’s still hesitation. According to Ad Age, only 45% of consumers say they support the use of AI in advertising. Thirty-six percent say they don’t. And the rest? They’re undecided.

So the question looming over marketers today isn’t can you use AI to make ads. It’s: Should you?

And deeper still: Will your audience care? Will your brand suffer? Will you get left behind if you don’t?

To get past the hype, I spoke to two people in the trenches:

• Barry Hott, a growth consultant (Hottgrowth.com) who has worked with brands like True Classic, The Perfect Jean, and Athletic Greens, and is already using AI to enhance performance creative.
• Matthew Gattozzi, founder of Goodo Studios, who shoots performance creative in person and has strong thoughts on the “creatives are cooked” narrative.

Here’s what we learned.

Performance is up—but so are the questions

"I’m already using it. We have AI hooks running in real ad campaigns right now," Barry told me. "And when it’s done right? They work."

He’s not the only one. According to the IAB, over half of ad buyers are now using generative AI in their workflows. The appeal is obvious: faster production, lower cost, higher volume.

That’s especially important given that 47% of sales are attributed to the creative itself, according to Nielsen. But Barry’s quick to clarify that this isn’t about replacing your creative team with a text prompt. It’s about scale.

"Even if I’m not using it for a full ad, I might use AI to generate the first three seconds. Just the hook. I can test five ideas today instead of waiting a week to coordinate shoots."

What he’s not doing: creating fake testimonials or having AI avatars pretend to be real customers.

"That’s not compliant, and it’s not ethical. But I am using AI to explain a problem, show a visual, or mock up a scenario I wouldn’t be able to shoot easily."

But is quality taking a hit?

It depends who you ask.

Matthew Gattozzi isn’t anti-AI, but he’s skeptical of the claims that AI will replace high-quality ad production.

"People say, 'We don’t need to hire a creative team anymore.' I just laugh," he said. "Most of them haven’t even used a camera."

Gattozzi’s studio creates performance-driven ads, not brand films, but they still shoot everything in person. For him, the value AI brings isn’t in skipping strategy, but in making it stronger.

"AI should make your process better, not cheaper for the sake of it. If your strategy sucks, no tool is going to fix that."

In his eyes, the real winners will be creatives who learn to wield AI, not the ones who avoid it… Or blindly trust it.

"You still need ideas. You still need taste. AI can help you get started. But it can’t tell you what’s worth saying."

His recommendation for brands? Focus less on replacing your agency and more on combining the right inputs. "If you're just using AI to skip the hard parts, you're not getting the value. Use it to brief better. Storyboard faster. Explore more."

The novelty is peaking. What comes next?

There’s no denying the engagement that AI ads are driving right now, especially on TikTok and Reels. But both Barry and Matthew agree: a lot of that engagement is still driven by novelty.

"You’re not watching because it’s a good ad. You’re watching because you’re like, ‘Wait, this was made with AI?’" Matthew explains.

Barry compares it to the NFT boom. A lot of what people are engaging with isn’t the message or story, it’s the format itself. That’s fine, for now. But novelty fades. Attention doesn’t last.

"At some point, people are going to scroll past AI Barbie just like they scroll past your generic product demo," Barry said. "You can’t rely on the gimmick forever."

Which brings us back to strategy.

What audiences actually want (hint: it’s not just polish)

If you think the solution is to make AI ads look more “real,” Barry disagrees.

"There are two kinds of AI content. One is trying to look human. The other leans into the weird. And both can work. You just have to make something people want to consume."

For some audiences, that might mean raw, lo-fi, TikTok-native videos. For others, it might mean stylized visuals, surreal storytelling, or humor that breaks the fourth wall.

But neither audience cares about how it was made.

"Most people don’t notice or don’t care if something was made with AI—at least right now, with what I’ve noticed," Barry said. "The only people losing their minds about this are ad people themselves."

WATCH: My interview with Barry

Tips for brands: When to use AI and when not to

So, how do you use AI well? Both Barry and Matthew offered a few soundbites:

1. Use AI to unblock creative production

Barry doesn’t believe in using AI to fully replace video teams. But if you’re stuck waiting for a shoot to be scheduled, AI might be the shortcut you need to keep momentum.

"Even if you’re not creating the full video with AI, you can use it for one scene, one transition, one visual metaphor," he said. "If that saves you from stalling out for a week, use it."

Use tools like VO and Cling to animate product shots, storyboard ideas, or rapidly mock up new hooks that would otherwise be costly or slow to shoot. Barry emphasized using AI-generated video in a modular way: plug it into your human-made content to speed up workflows without sacrificing trust or taste.

2. Start small: Hooks, B-roll, transitions

Matthew echoed this point. He recommends that brands focus on parts of their ads where AI can genuinely accelerate production without compromising on quality.

"You don’t need to AI-generate your entire ad. But can it help you visualize a product demo? Can it animate a customer quote or create transitions between scenes? That’s where it shines."

For example, he often sees brands stuck re-filming content because the tone or pacing of the original wasn’t quite right. With AI, you can test voiceover tone, pacing, inflection, and emotion—all before committing to a full shoot.

3. Don’t use it to fake authenticity

Both experts were firm on this point: using AI to create fake testimonials or avatar-based "customers" crosses a line.

"It’s not just a legal risk. It’s a trust risk," Barry said. "Your audience might not notice the first time, but if they do—and they will—it could damage more than just campaign performance."

Instead, use AI to enhance real stories. Layer AI-generated content behind human creators. Supplement real UGC with animated explanations or added context. Keep the line between fiction and reality crystal clear.

4. Use AI to stress-test creative before big production investments

Barry calls this "pre-shoot validation."

"One of the most expensive mistakes a brand can make is going all-in on a concept that hasn’t been tested," he said. "AI lets you storyboard faster, write scripts faster, and get feedback sooner."

Before you book a $30K shoot, generate a scrappy version of the idea in AI. Show it to your team. Test it with a small ad budget. Iterate.

"AI helps you avoid the sunk cost fallacy," Barry added. "You can walk away from a bad idea before it becomes an expensive one."

5. Feed AI the context it needs to perform well

Barry noted that the worst AI outputs happen when marketers treat it like a magic button.

"AI is only as smart as your prompt," he said. "If you want great outputs, give it great inputs. Brand voice, audience details, product features—pack your prompts with context."

Matthew agreed. "If you can’t articulate your strategy clearly, you won’t be able to get good results from AI either. It forces you to be sharper, which is a good thing."

6. Know your platform, know your buyer

What works on TikTok won’t necessarily work on YouTube. What delights a Gen Z skincare shopper might alienate a B2B software buyer.

AI gives you the ability to produce at volume—but that doesn’t mean you should spray and pray.

"Some brands can lean into the AI aesthetic. Others shouldn’t touch it for awareness campaigns," Barry said. "Know who you're talking to and test in a low-risk way first."

7. Use AI to go faster, not to get lazy

"This isn’t a shortcut to avoid creative work," Matthew said. "It’s a tool to make your ideas stronger, faster."

Use it to:

• Remove the blank page problem
• Turn good ideas into scripts faster
• Visualize edits and transitions
• Refine tone and message

But never use it as a reason to stop thinking critically.

"If your entire plan is: 'Let’s see what the AI spits out,' you don’t have a plan." 

8. And if you're scared of falling behind—start small

You don’t need to go all-in. But you do need to start.

"The people who are cooked are the ones sitting this out," Barry said. "Even if you're just animating stills or playing with transitions, dip a toe in. The tools are evolving fast."

AI ad swipe file: Heinz’s "Even AI Knows It’s Heinz"

In a now-viral spot, Heinz asked DALL-E 2 to generate images based on prompts like "ketchup," "ketchup in outer space," or "ketchup street art."

The resulting images ranged from gothic stained-glass renderings to abstract ketchup bottles—but each one, whether intentionally or not, closely resembled Heinz’s iconic branding.

The ad flipped the AI trend into a story about brand recognition. It ends with the line: "Even Artificial Intelligence Knows 'Ketchup' Looks Like Heinz."

Not only did the brand encourage viewers to create their own AI-generated ketchup images, it sparked conversation without relying on deepfakes or gimmicks. It used AI to prove brand equity.

These images even found their way to the metaverse… 

According to Noor Salam Khan, an AI marketing leader, here are some results from the campaign:

• Over 800 million earned impressions globally
• Their media investment was exceeded by over 2500%

“This campaign only worked because Heinz is ketchup. AI simply proved it. What's the unshakeable truth about your brand?” Noor says in his post.

This is the kind of AI ad execution worth studying: audience-first, culturally timely, and built around a concept that strengthens the brand rather than distracts from it.

P.S. The environmental cost no one’s talking about

Here’s the uncomfortable truth: AI ads may be fast, cheap, and scalable—but they’re not free.

Literally.

A study of 88 AI models found that one single image generation can use the energy equivalent of half a smartphone charge. The most carbon-intensive models? Equivalent to driving a gas-powered car for 4.1 miles—for every 1,000 outputs.

And if you’re generating hundreds or thousands of variations to test creative? 

That footprint adds up fast.

"Please don’t write this without mentioning the environmental impact," Barry said. "Even if we can’t stop it, we should at least acknowledge it."

So… are creatives actually cooked?

If you work in ads, the answer is: No—but your complacency might be.

"The people who aren’t learning this stuff? They’re already cooked," Barry said.

Matthew agrees. "The iPhone made it easier than ever to shoot a video. If you didn’t do it then, why do you think you’ll do it now with AI? You still have to want to create."

More tools doesn’t mean more great content. It just means the excuses are gone.

And if you're scared of AI replacing you?

"I have the same tools you do," Matthew said. "But I also have taste, strategy, and 10 years of experience making videos. You want to replace me? Good luck."

Final Thought

AI is changing advertising. That part’s not up for debate.

But the best marketers won’t be the ones who abandon craft. They’ll be the ones who learn how to blend it—at scale.

Use AI to move faster. To test more. To explore ideas you wouldn’t otherwise afford.

Just don’t forget the one thing AI still can’t do (at least right now): understand your audience and strategize like a pro.

‍