By Craig Benner, Founder & CEO, Accretive
Super Bowl LX is over, and one thing is unmistakable: Last year’s ads talked about AI, while this year’s ads increasingly used it. In fact, 23% of Super Bowl commercials this year either promoted AI or used it in some way to create their ads.
Here’s my heretical take: AI doesn’t always make Super Bowl ads better. In many cases, it’s making them easier to ignore.
AI is excellent at producing competent outputs at scale, and that’s exactly the trap. When everyone has access to similar models, similar prompts, similar “best practices,” and similar guardrails, creative converges. You can almost feel it: ads that look expensive, sound polished, and vanish from memory the second the screen fades to black.
NielsenIQ’s research gets at why: Consumers often perceive AI-generated ads as more “annoying,” “boring,” and “confusing,” and even “high-quality” AI ads can trigger weaker memory activation than traditional creative.
That matters because the Super Bowl isn’t a click campaign. When you buy a Super Bowl advertising slot, you’re buying the opportunity to be a part of a cultural memory.
Eeriness kills the buy-in
If you want a live-fire example of what audiences mean by “uncanny,” this year handed advertisers one on a platter: Svedka’s mostly AI-generated spot was widely positioned as a milestone for generative video on the Big Game stage, but it also drew plenty of “nightmare fuel” reactions.
The lesson isn’t “never use generative video.” It’s that synthetic media gets judged on more than production value. It gets judged on human intent. When the creative is meant to feel emotional, intimate, or aspirational, the tolerance for “close enough” drops fast.
Research backs up the intuition, showing that people can perceive AI-written communications as less authentic, which can shape downstream reactions. And separate research on AI-generated advertising suggests that perceived eeriness reduces acceptance.
Translation: The more your ad feels like it was generated by AI, the more you’re forcing audiences to decide whether they trust you before they even decide whether they like you.
If the use case is fuzzy, people smell it
You don’t have to be a creative purist to notice that some AI advertising struggles with the most basic job of an ad, which is explaining the payoff in plain language.
This year’s AI-heavy lineup made that contrast obvious. Some brands landed the plane by pairing AI with a clear, everyday benefit. Kellogg School Super Bowl Advertising Review, for example, praised Google’s Gemini spot for combining emotional storytelling with a clear, natural use case. Meanwhile, other AI-forward ads (like Artlist’s) drew failing grades in post-game scoring, an indicator that “AI” as a headline isn’t automatically a hook.
That matters because the Super Bowl is high-attention but low-patience. Viewers will follow absurdity if the premise is clear, but if the use case doesn’t map to a recognizable human problem or tell a compelling story, it reads as “tech for tech’s sake,” and that can backfire. CivicScience found that only 6% of viewers said Super Bowl AI ads made them more likely to use AI, while one-quarter said they were less likely.
4 advertising rules to avoid AI-driven sameness
Even if the Super Bowl is behind us, the creative dynamics it exposes are year-round. These are implementable guardrails for teams trying to use AI without sliding into sameness or uncanny territory.
- Lock in your brand voice before you prompt.
Write a one-page “voice sheet” (3 traits, 5 dos and don’ts, and 10 approved phrases). Use AI only to create variants inside that guardrail, never to generate the voice itself. - Ban synthetic sentiment.
If the goal is warmth, nostalgia, or inspiration, default to real humans, real footage, and real stories. If you need to use generative elements, use them for background or utility, not faces, “memories,” or emotional beats. - Lead with the human problem.
Before you ship creative, force a one-line “job to be done” statement: “This helps you ___ without ___.” If you can’t say it that simply, the idea’s not ready. - Test for recall.
Run a 24-hour “name the brand and what it was about” check with a small internal or customer panel. If people can’t repeat the premise in one sentence and connect it back to you, keep iterating.
In 2026, production isn’t a scarce resource, but perspective can be.
AI can help you make more ads, but it cannot make distinctive ads on its own. And on Super Bowl Sunday, “more” is exactly how you become wallpaper. If you want to win, don’t ask, “How do we use AI?” Instead, ask: “What could only we say?” Then, use AI to scale that clarity, not replace it.
About the Author
Mike Tretinjak is VP of enterprise strategy at Intero Digital. Mike has over 20 years of experience in digital marketing, and he’s known for helping brands stay ahead as SEO, paid media, GEO, and AI-driven strategies evolve at breakneck speed. He leads data-driven, innovation-first programs, and he’s always up for a conversation about AI, travel, golf, or the Kentucky Wildcats.
