Study Shows Brands Benefit from Ads Next to AI-Generated Content

By Charles Ryder

For the past couple of years, brand safety conversations have tended to land in the same place whenever AI-generated content comes up: avoid it. The working assumption has been that if a video is AI-made, it’s automatically risky—more “AI slop” that could embarrass advertisers through bad adjacency.

A new, first-of-its-kind study suggests that instinct may be costing brands performance.

In research released March 2, 2026, brand suitability firm Zefr and Omnicom’s OM Media Trials tested how ads perform when placed next to different categories of AI-generated video. The results were more nuanced—and more encouraging—than the usual panic cycle suggests.

The “AI Slop” Era Created a Blanket Fear

From mid-2023 through late-2025, AI video tools improved quickly, and platforms flooded with everything from clever experiments to outright spam. “AI slop” became shorthand for low-quality, deceptive, mass-produced content built for clicks and ad revenue—uncomfortably similar to the made-for-advertising (MFA) mess the industry wrestled with in the 2010s.

Advertisers responded the way they often do under uncertainty: broad exclusions, rigid keyword blocks, and internal policies that treated “AI” as one high-risk bucket. It’s understandable when you’re protecting brand trust at scale. It’s also a blunt instrument.

What the Zefr + OM Media Trials Study Actually Tested

This study didn’t ask a fuzzy question like “Is AI content good or bad?” It broke AI video into eight types and measured how adjacency affected real brand metrics.

Here’s the core setup:

• Around 5,000 consumers in the U.S. and Canada were surveyed.
• Participants viewed ads adjacent to multiple AI-generated video categories.
• Researchers measured outcomes including ad recall, favorability, trust, and purchase intent.

The content types ranged from creative, clearly entertainment-driven formats (satire, youth humor, artistic content) to environments brands have good reason to worry about (spam-like content, misinformation, and sensitive or deceptive depictions).

That distinction—which kind of AI content—is the whole story.

The Big Finding: Not All AI Adjacency Hurts Brands

The headline takeaway is simple: ads placed next to certain AI-generated videos can perform better than expected.

According to the reported findings, adjacency to more creative AI content types—like satire, youth-oriented humor, and artistic AI videos—can lift:

• Ad recall
• Positive associations, including the brand being perceived as “innovative”

That tracks with how people actually consume feeds. Entertainment and novelty earn attention. If the surrounding content is engaging—and not trying to trick anyone—the ad can benefit from that attention carryover.

AI doesn’t automatically degrade context. In some placements, it can improve it.

The Bad News: Unclear or Spammy AI Contexts Still Damage Trust

This wasn’t a free pass for AI inventory. The study also found meaningful downside when AI content is:

• Spam-like
• Misinformative
• Unclear in origin or intent, especially when viewers can’t tell what’s real and what’s synthetic

Those environments reduced favorability, trust, and purchase intent, with some categories—especially finance—appearing more sensitive to reputational harm.

One stat stands out: 81% of respondents said at least one AI content type is inappropriate for brands to advertise alongside, and Canadian respondents were more sensitive overall. Suitability isn’t just category-specific; it can be market-specific too.

Why Labels and Transparency Matter More Than Ever

The research also underlines how imperfect detection and disclosure still are. People frequently misidentify content—sometimes assuming human-made content is AI-generated, sometimes missing that AI was used at all.

That confusion has a cost. When audiences aren’t sure what they’re watching, confidence drops—and brands get caught in the uncertainty.

Transparency helps. The study indicates that 41% of consumers feel better about brands when AI-generated content is clearly labeled. That’s not just a policy checkbox; it’s a performance lever. Labels reduce ambiguity, and ambiguity is where trust erodes fast.

The Real Shift: From “Block AI” to “Control AI”

The strategic takeaway is straightforward: treating AI as one giant risk bucket is becoming a losing strategy.

Zefr’s position—reflected in the study—is that brands need granular controls, not blanket avoidance. In practice, that means suitability frameworks that can:

• Align with high-quality AI creativity (satire, art, entertainment)
• Limit categories where context is murky or intent is unclear
• Avoid clearly harmful segments like spam and misinformation

It’s the same maturity curve we’ve seen before. The winners aren’t the brands that hide from the open internet—they’re the ones that learn to navigate it with better tools and tighter controls.

What This Means for Advertisers Right Now

If you’re running paid media in 2026, this study makes three realities hard to ignore:

1. AI inventory will be unavoidable at scale.
   Whether Gartner’s “90% by 2030” prediction lands exactly or not, AI-assisted creation is becoming the default.
2. The premium opportunity is in the “good AI” environments.
   Creative AI content isn’t synonymous with slop. Some of it drives attention, recall, and an innovation halo.
3. Your competitive edge will come from nuance.
   Brand suitability is moving beyond “avoid risk at all costs.” The new advantage is segmentation, labeling, and adjacency rules tailored to your category and markets.

FAQ

Does advertising next to AI-generated content always hurt brand perception?

No. The study found that certain creative AI contexts (satire, youth humor, artistic AI) can improve outcomes like ad recall and perceptions of innovation, while spammy or deceptive AI contexts tend to harm trust and purchase intent.

Which AI content environments are riskiest for brands?

Spam-like, misinformative, or unclear AI content—especially when viewers can’t tell what’s real—showed the strongest negative impact on favorability, trust, and purchase intent.

Do consumers care whether AI content is labeled?

Yes. The study reports that 41% of consumers feel better about brands when AI-generated content is clearly labeled, suggesting disclosure can reduce uncertainty and improve sentiment.

Conclusion: The Opportunity Is Real—If You Stop Treating AI as a Monolith

This study isn’t a reason to sprint into every AI-generated feed. It’s a clear signal that the fear-based default—block anything AI—is too simplistic for where media is heading.

AI content is a spectrum. Some of it is artistic, satirical, and transparently “in on the joke.” Some of it is spam, manipulation, or misinformation dressed up as entertainment. Brands don’t need a bigger hammer. They need better controls, clearer definitions, and smarter decision-making that protects trust without sacrificing reach.

If you’re building a modern framework for evaluating AI content and the ad ecosystems around it, use a toolset that treats context with the nuance it deserves. AIuthority can help you track the conversation, understand the categories, and stay ahead as AI-generated media becomes a normal part of the advertising landscape.