BusinessWire
AI‑generated videos, images, and text have moved from novelty to mainstream across TikTok, Instagram, YouTube Shorts and other short‑form platforms. Brands that buy inventory on these feeds now confront a dilemma: should they avoid any AI‑powered content, or can they selectively leverage the formats that resonate with consumers? Until now, most brand‑safety research has focused on how advertisers programmatic use AI in creative assets, leaving a blind spot around the impact of the surrounding AI‑driven environment.
By measuring ad recall, perception of innovation, favorability, trust and purchase intent in direct response to the type of AI content that appears next to an ad, Zefr and OM Media Trials provide the first empirical evidence to inform that decision‑making.
Scope and methodology
The study was executed by OM Media Trials, a research unit within Omnicom Media that specializes in custom investigations for media owners and ad‑tech platforms. Participants from both the U.S. and Canada were exposed to a series of ad placements adjacent to a curated set of AI‑generated assets. The assets were grouped into sub‑categories such as:
- Satirical or humorous AI depictions of youth
- Creative‑expression pieces generated by AI tools
- Spam‑like or deceptive AI videos
- AI‑generated content featuring public figures or sexualized imagery
The researchers tracked standard brand‑impact metrics—ad recall, perceived innovation, favorability, trust and purchase intent—while also probing respondents’ ability to discern whether a piece of content was human‑made or AI‑generated.
Key takeaways
Not all AI is “slop”
The headline finding is that AI content is not a monolith. In categories that emphasized satire, humor or artistic expression, ads placed next to the AI material actually outperformed baseline measurements. Respondents reported higher recall and a stronger impression that the brand was innovative. This suggests that certain AI environments can amplify a campaign’s reach rather than dilute it.
Spam‑like AI hurts brand perception
Conversely, when the adjacent AI content resembled low‑quality, misleading or spammy material, the same brand metrics dipped. The study links these negative outcomes to consumer uncertainty—viewers struggled to tell whether the surrounding media was authentic or fabricated, which eroded trust and purchase intent.
Consumers draw firm lines
A striking 81 % of participants indicated that at least one type of AI‑generated content should be off‑limits for brand adjacency. The data underscores a growing consumer appetite for brand‑safe environments, even as AI becomes more ubiquitous.
Geographic nuance
Canadian respondents proved more sensitive to AI‑adjacency risks than their U.S. counterparts. The heightened caution in Canada could reflect differing regulatory climates or cultural attitudes toward AI and digital media.
Transparency improves outcomes
Clarity emerged as a powerful lever. When AI‑generated content was clearly labeled, 41 % of respondents said their perception of the brand improved. Trusted creators and explicit disclosure of AI tools further bolstered brand sentiment. In contrast, ambiguous or undisclosed AI content triggered declines across favorability, trust and purchase intent.
Voices from the field
“*AI content is rapidly becoming unavoidable for advertisers, but treating all AI as a single risk category is both inaccurate and limiting*,” said Jon Morra, Chief AI Officer at Zefr. “*This research shows that some AI environments can drive positive brand outcomes, while others introduce real brand risk. The difference lies in the type of AI content and how it aligns with brand values.*”
Adding a forward‑looking perspective, Kara Manatt, EVP of Intelligence Solutions at OM Media Trials, remarked, “*AI content is only going to become more prevalent in the months and years ahead.* *The solution is not to shut off an entire category of content, but to give brands the control and intelligence to align with the right AI environments, and avoid the ones that create risk.*”
Industry context
The findings dovetail with a broader shift in ad tech toward granular suitability controls. Platforms such as Meta, TikTok and YouTube have introduced AI‑labeling initiatives, but adoption remains uneven. The study’s emphasis on clear labeling aligns with emerging regulatory guidance in the EU and U.K., where transparency around synthetic media is increasingly mandated.
Moreover, the data reinforces the business case for AI‑aware brand‑suitability engines. Companies that can programmatically filter out low‑quality AI assets while allowing high‑engagement, brand‑safe AI content stand to capture incremental lift without sacrificing safety.
Practical implications for marketers
- Refine suitability policies – Instead of a blanket ban on AI, marketers should categorize AI content by quality and relevance, mirroring the sub‑categories used in the study.
- Leverage transparent labeling – Partner with platforms that provide clear AI disclosures, or add in‑house labels where possible, to capture the 41 % uplift observed.
- Monitor regional sentiment – Canadian campaigns may require stricter AI‑adjacency filters than U.S. efforts, given the heightened sensitivity.
- Collaborate with trusted creators – Aligning with influencers who openly disclose AI tools can mitigate consumer uncertainty.
How Zefr can help
Brands interested in applying the study’s insights can work directly with Zefr to set bespoke AI suitability thresholds. Zefr’s technology enables customized controls that let advertisers align with, limit or avoid specific AI content categories based on their own risk appetites and brand values.
The full report, complete with methodology details and raw data, is available for download at https://zefr.com/ai-slop-or-not.
Looking ahead
As generative AI tools become embedded in everyday creator workflows, the volume of AI‑generated video and imagery is expected to rise sharply. The study’s early evidence suggests that the market will soon differentiate between “AI‑friendly” and “AI‑risky” environments. Future research will likely explore longitudinal effects—whether the positive lift from high‑quality AI persists over time, and how evolving consumer awareness of synthetic media reshapes brand‑safety standards.
For now, the takeaway for ad tech professionals is clear: not all AI is created equal. Strategic, data‑backed filtering combined with transparent labeling can turn AI from a perceived liability into a source of incremental brand value.
Get in touch with our Adtech experts
