Coca-Cola has launched its 2026 U.S. holiday campaign with a creative choice that cuts against the last two seasons: its new “Enjoy Joy” spots move away from generative AI. The films show families and friends sharing familiar holiday moments, and the campaign spans television, streaming, digital and social. The move is worth watching, but it is not proof that AI advertising has failed. It is a reminder that production technology is a strategic choice, not a creative strategy by itself. Marketing Dive’s report describes the shift, while Coca-Cola’s official announcement describes the campaign without mentioning AI or explaining the production decision.

Read the change narrowly, not as a retreat

The “Enjoy Joy” spots were created by WPP Open X, led by Ogilvy, directed by Paul Dektor and produced by Mayan Productions, according to Marketing Dive. They come in 30-, 15- and six-second cuts, with versions tailored for general-market and Hispanic audiences. The brand says they will run across linear TV, digital, streaming and social platforms. That is a broad, multi-format campaign—not a rejection of every use of AI across Coca-Cola’s marketing.

That distinction matters. Marketing Dive reports that Coca-Cola used generative AI in holiday advertising during the prior two years, and that the earlier ads exceeded internal benchmarks while proving divisive with consumers. But the company has not publicly said that audience criticism caused this year’s production decision. Its official campaign release emphasizes connection and togetherness, not an AI policy or a performance comparison. Marketers should treat this as a signal about one campaign’s creative fit, not a company-wide verdict on AI.

Why “top-tested” can coexist with divided reactions

In 2025, Coca-Cola executive Islam ElDessouky called an AI holiday ad “one of our top-tested ads in history” and said it scored highly on measures including brand association and conversion to transaction. Those are company-reported claims: the underlying scores, research design and comparison group were not disclosed in the interview. Separately, Marketing Dive described the campaigns as divisive with consumers, without quantifying the size or representativeness of that response. Both signals can be true because a test metric, an online conversation and long-term brand meaning are different things. The executive’s comments and the limits of the disclosed data should be read together.

The practical lesson is not to optimize for applause or backlash alone. Before approving AI-generated creative, decide what the work must accomplish: attention, brand linkage, emotional response, conversion, speed or lower production cost. A campaign can win one measure and still create an unwanted trade-off on another. Report those outcomes separately rather than compressing them into a single “AI worked” headline.

What the evidence says about AI-made advertising

Broader evidence argues against a simple yes-or-no rule. Kantar’s analysis of hundreds of ads that used generative AI found that more than 40% of ads where AI was integrated seamlessly landed in the top tier for branded cut-through. Ads with distracting or unnatural visuals—often cases where AI use was obvious—tended to perform worse. Kantar’s conclusion: “The key isn’t whether you use GenAI, but how you use it.” This is an analysis of ads in Kantar’s database, not a controlled test proving that AI itself caused stronger or weaker results.

That nuance changes the brief. If a visual requires believable human expression, emotional subtlety or a strong sense of place, those are demanding creative constraints—not automatic reasons to use or avoid AI. Generative tools may help teams explore concepts, create storyboards, adapt versions or speed up iteration. Human direction, brand judgment and quality control remain essential, especially when the campaign depends on warmth and authenticity.

A practical decision process for marketing teams

  • Start with the audience and objective. Name the emotion, behavior and brand memory the work needs to create before choosing production tools.
  • Use AI where it adds leverage. Test it in ideation, previsualization, localization or variant production when it improves speed or range without weakening the central idea.
  • Compare like with like. Test AI-assisted and conventional executions with the same audience, placement and success criteria. Track brand linkage and intended emotion alongside response and cost.
  • Set a human review gate. Inspect faces, motion, cultural details, product accuracy, rights and disclosure needs. If an AI artifact distracts from the story, fix it or choose another approach.
  • Keep the evidence honest. Separate measured results from internal claims, social sentiment and assumptions about why a campaign changed.

Coca-Cola’s 2026 choice is most useful as a case study in disciplined selection: use the method that serves the story, then prove it with measurement. Real Internet Sales helps businesses build practical AI and digital marketing programs around clear goals, quality execution and accountable results. Call 803-708-5514 or visit realinternetsales.com to discuss your strategy.