AI-Generated Ads: The Quality Gap Between Hype and Reality

A Disconnect Between Adoption and Quality

83% of ad executives say their company has deployed AI in the creative process in 2026, up from 60% in 2024. Meanwhile, consumer sentiment toward AI-generated advertising continues to decline, with the perception gap between advertiser optimism and consumer skepticism widening.

Numbers tell a story of misaligned incentives. We're racing to adopt AI tools because they're fast and cheap, not because they're producing better ads. And the evidence suggests that speed and cost savings don't automatically translate to creative quality or business results.

Performance Data: Mixed at Best

Strongest evidence for AI-generated ad effectiveness comes from an MIT study of 21,000 consumers. AI-generated personalized video ads delivered 9.4% higher click-through rates than personalized image ads and 6.5% higher CTR than generic videos. That's meaningful, but it's also narrow, one format, one metric, one study.

Click-through rates tell us about immediate response, not about the kind of mental availability that drives long-term brand growth. As Patrick Gilbert explores in Never Always, Never Never, metrics that matter most for sustainable growth often can't be measured in the short term.

MIT's researchers were careful to flag limitations of their findings: open questions around novelty effects, privacy concerns, and quality at scale. They explicitly warned that efficiency gains shouldn't come at the expense of human judgment or consumer trust. That warning seems prescient given what we're seeing in the broader market.

Trust Problems Are Getting Worse

71% of Gen Z and Millennial consumers said they believe they have seen an ad created using AI, up from 54% in 2024. More exposure should theoretically lead to greater acceptance. Instead, it's driving more skepticism.

More than half of consumer respondents said advertisers should disclose if an ad is 100% AI-generated, uses AI video, or includes AI images. Nearly half wanted disclosure for AI voices or virtual characters. These aren't fringe opinions, they're mainstream consumer expectations that most brands are ignoring.

A University of Kansas study examined 1,375 AI-generated programmatic ads and found that only about half were clearly labeled as such. Transparency problems aren't just about consumer preference; they're about basic honesty in advertising.

Consumer trust drives emotional advertising effectiveness. When consumers suspect they're being misled about how an ad was created, it undermines emotional connection that makes advertising memorable and persuasive.

Why AI Struggles With Emotional Depth

Best advertising doesn't just communicate features or benefits, it makes people feel something that connects back to the brand. AI-generated creative hits its biggest limitation here.

Gilbert writes about Marvin Waldman's work on Campbell's Chunky Soup, which repositioned soup as comfort food by tapping into men's emotional connections to their mothers. Campbell's "Mama's Boy" campaign worked because it understood a deeper human truth: soup equals mom, and there's nothing more emotional than that relationship.

AI can generate variations on existing creative themes, but it struggles with the kind of cultural insight that drives breakthrough campaigns. It can optimize for engagement metrics, but it can't intuit emotional undercurrents that make advertising stick in memory.

That's why distinctive brand assets built through human creativity like the Safelite jingle or the Duolingo owl remain more powerful than algorithmically optimized creative. They're rooted in understanding how people actually think and feel, not just how they click.

Infrastructure Problems

AI creative tools suffer from what Gilbert describes as "AI wrapper" problems in Never Always, Never Never. They look impressive in demos but fall apart when you rely on them for real work.

The issue isn't the AI itself, it's the infrastructure underneath. As Gilbert's team discovered while building their strategic analysis tools, effective AI applications require deterministic backend systems that ensure reliability. Most AI creative tools skip this foundation, resulting in outputs that are fluent but unreliable.

Broader challenges in AI adoption mirror what Gilbert explores through the AI double helix framework. First strand, using AI for efficiency, is relatively straightforward. Second strand, using AI to create new value, requires much deeper infrastructure and strategic thinking.

Agencies and brands are still operating in the first strand, using AI to produce more creative faster rather than to solve fundamental creative challenges. That approach might reduce costs, but it doesn't necessarily improve quality.

Real Opportunity: AI as Creative Infrastructure

Promising applications of AI in creative aren't about replacing human creativity but about filling resource gaps that prevent good creative from being executed well.

Gilbert describes how AdVenture Media uses AI to maintain consistent blog content production, not because AI writes better than humans, but because their marketing team lacked the bandwidth to publish consistently. AI handles routine work while humans focus on strategy and quality control.

A more sustainable model for AI in creative emerges: using automation to handle mechanical aspects of production while preserving human judgment for strategic and emotional decisions.

That could mean AI-generated variations of proven creative concepts, automated adaptation of successful campaigns across formats, or dynamic optimization of messaging while maintaining consistent brand voice. Key is using AI to amplify human creativity rather than replace it.

What Actually Works

Evidence suggests AI-generated ads can be effective when they're used strategically within broader creative frameworks developed by humans.

Personalized video, as shown in the MIT study, benefits from AI's ability to customize content at scale. But underlying creative concept, story, emotional hook, and brand connection still needs human insight.

Disclosure, contrary to industry fears, may actually improve performance. IAB's research indicates that clear labeling can improve consumer attitudes and even increase purchase likelihood. Transparency becomes a competitive advantage rather than just a compliance requirement.

Brands that will succeed with AI creative are those that treat it as infrastructure rather than strategy. They'll use AI to execute human-developed creative concepts more efficiently and at greater scale, while preserving emotional depth and cultural insight that makes advertising memorable.

Bottom Line

AI-generated ads aren't inherently good or bad, they're tools that amplify whatever creative strategy you bring to them. If your underlying approach to advertising is mechanistic and short-term focused, AI will help you produce more mediocre creative faster. If your approach is grounded in human insight and emotional understanding, AI can help you execute that vision more effectively.

Current gap between adoption rates and consumer sentiment suggests most brands are optimizing for the wrong metrics. They're measuring efficiency over effectiveness, speed over emotional impact, cost savings over long-term brand building.

As Gilbert argues throughout Never Always, Never Never, strategic advantage belongs to marketers who understand that AI's real value lies not in replacing human judgment but in freeing human creativity from resource constraints. Question isn't whether AI can generate good ads, it's whether you have a creative strategy worth amplifying.