See if we're a fit

AI creative

Do AI Ads Actually Work? An Honest Look at the Evidence

Gabe Hutcheon · · 6 min read

AI-generated ads match or beat human-made ads for lower-AOV, product-driven creative and are excellent at producing test volume cheaply. They underperform on high-AOV, emotional and storytelling briefs, where human-made ads still convert better. The winning move is a hybrid, not all-in either way.

The AI-ads conversation is split between tool vendors saying AI wins everything and skeptics saying it never works. Neither is true. The honest answer depends on what you are selling and what you are using AI for. Here is where AI ads genuinely work and where they still lose.

What the independent research says

Set aside the vendor marketing. Independent research (a study by Ipsos with Syracuse University) found two useful things: viewers largely cannot reliably tell AI-made ads from human-made ones, and yet human-made ads tend to over-index on short-term sales impact while AI-made ads under-index. In plain terms: AI ads are good enough to pass, but not always good enough to perform, especially for considered purchases. Treat any single statistic as directional and weigh it against your own results.

Where AI ads win

  • Test volume. AI can produce many variations for a fraction of the cost and time, which means more shots at finding a winner. This is its biggest real advantage.
  • Lower-AOV, product-driven creative. For straightforward, impulse-friendly products, AI ads can match or beat human-made ones on ROAS.
  • Prototyping. Use AI to find which angles and hooks have legs before committing a production budget to them.

Where humans still win

  • High-AOV, considered purchases. The more thought a purchase requires, the more trust and nuance matter, and the more human-made creative outperforms.
  • Emotional and storytelling briefs. AI is weakest exactly where emotion carries the ad.
  • Credibility-dependent products. Where authenticity is the whole sell, a real person still converts better.

The hybrid workflow

The brands getting real value do not choose. They use AI to prototype and test concepts at volume, identify the winners cheaply, then scale those winners with human-made or human-led creative, especially for higher-AOV or emotional products. AI widens the top of the testing funnel; humans win the bottom. This maps cleanly onto the production mix in how many ad creatives you need per month.

The takeaway

"Do AI ads work?" is the wrong question. The right one is "for what?". For volume and lower-AOV product creative, yes. For high-AOV and emotional work, not yet. Use it as a testing accelerant, not a replacement for judgement or for human creative where it counts. For how AI UGC fits alongside real creators, see whether UGC ads still work, or book a free creative audit to work out where AI fits in your account.

Frequently asked questions

Do AI ads actually work?
Yes, in the right place. AI-generated ads match or beat human-made ads for lower-AOV, product-driven creative and excel at producing test volume cheaply. They underperform on high-AOV, emotional and storytelling briefs, where human-made ads still convert better.
Can people tell if an ad is AI?
Mostly not. Independent research (Ipsos with Syracuse University) found viewers struggle to reliably distinguish AI-made from human-made ads. The bigger issue is not detection, it is that AI ads tend to underperform on short-term sales for considered purchases.
Should I replace my creative team with AI?
No. The best results come from a hybrid: use AI to prototype and test concepts cheaply at volume, then scale the winners with human-made or human-led creative, especially for high-AOV or emotional products.

Think creative is your bottleneck?

We brief from $250M+ in tracked ad spend and put first drafts in your hands in 48 hours. Book a free creative audit and we will show you where your account is leaking.

Book a free creative audit