When AI Misses the Mark: What CMOs Need to Watch For
- 21 hours ago
- 4 min read
AI adoption has grown quickly among marketing teams this past year. Many leaders now leverage it in their daily workflows, from early drafts and research, to analytics and creative exploration. But as marketers expand their AI use cases, the gaps in its capabilities are also becoming clearer. AI is not uniformly useful across all marketing activities. While it performs well in structured areas with high-volume inputs and measurable outcomes, it falters when the work depends on nuance or judgment.
Across our work at 621, one insight keeps proving itself: AI can effectively predict patterns, but it does not necessarily understand them. Gartner reported in 2024 that more than 80% of AI failures in marketing were caused by poor or incomplete data. When the inputs are off, the output is often suspect. Further, when the output requires interpretation, human judgment still carries the weight.
In our earlier blogs, we focused on where AI strengthens demand gen, analytics, revenue operations, and creative workflows. In this piece, we focus on the other half of the equation: the areas where AI falls short, where missteps have already played out in public, and where CMOs should establish guardrails before broader integration.
1. When AI Undermines Brand Strategy
Brand identity depends on interpretation and understanding of what customers value, cultural trends, and lived experience. AI can cluster and summarize. However, it cannot define meaning.
Example: A Retail Brand’s AI-Generated Campaign
What happened: A major apparel brand introduced AI-generated models into its e-commerce experience, positioning it as a way to increase diversity and improve personalization. The public reaction was immediate and overwhelmingly negative. Customers and industry leaders called the move inauthentic and a shortcut that avoided investing in real representation.
Why it happened: Audiences saw a mismatch between the company’s stated values and the execution. Even with good intentions, AI could not replace the trust that comes from real human representation in the process of framing the strategic direction of the brand.
Lesson: Authenticity is non-negotiable in brand strategy. Research shows that 86% of consumers say authenticity is a key factor in deciding which brands they support, and AI-generated stand-ins can be perceived as shortcuts. Human judgement must set the standard.
2. When AI Falls Short on Creative Originality
The brand ideas that endure - like “Just Do It” or “Got Milk” - were not statistical outputs. They were human interpretations of cultural moments, customer motivations, and emotional truths.
When brands lean on AI for this breakthrough, the results are often lackluster.
Example: A Legacy Brand’s AI-Generated Origin Film
What happened: A legacy brand released an AI-generated video intended to evoke the founder’s origin story. Instead of nostalgia, viewers fixated on visual inconsistencies: characters that changed appearance, unnatural movements, and distorted background details. As the campaign went viral for the wrong reasons, social sentiment swung negative.
Why it happened: AI’s generative strength lies in producing variations, not preserving continuity. Even with extensive prompting, the system struggled to maintain stable character identity, natural motion, and believable emotional cues across the full narrative arc.
Lesson: The campaign saw positive sentiment collapse 75% and negative sentiment climb above 50%. When creative feels off, audiences feel it immediately. CMOs should use AI for iteration, but keep humans in charge of the overall narrative.
3. When AI Overreaches in Leadership and Strategy
Effective marketing strategy requires savvy judgment, tradeoffs, and an understanding of how decisions land with real people. AI can support strategic work by surfacing inputs, but it cannot lead it.
Example: A Financial Platform’s AI-Driven Support Failure
What happened: A major tax platform introduced an AI assistant to answer user questions during filing. Independent evaluations found the system often delivered misleading or incomplete guidance, including confident responses that would have resulted in incorrect returns.
Why it happened: Generative AI surfaced statistically likely answers rather than applying regulatory logic or contextual judgment. Tax preparation relies on interpreting edge cases and nuanced definitions (areas where general-purpose models struggle without expert-curated data).
Lesson: AI’s accuracy breaks down when it has to interpret nuanced regulations. Public testing found one AI tax assistant produced wrong or misleading answers 50% of the time, underscoring how quickly trust erodes when AI oversteps into judgment-heavy work.
What CMOs Should Do Now
We’ve all seen the high-profile AI misfires that dominated headlines. But for every public stumble, there are countless quieter setbacks inside organizations—AI pilots that never gain traction, tools teams won’t adopt, and “smart” campaigns that fall well short of expectations. The truth is, AI isn’t a magic bullet. It can accelerate great marketing, but it can’t yet fix foundational gaps or replace strategic clarity.
AI will continue to evolve and improve, so it’s important to stay on top of the latest tools and trends. At the moment, marketing is still operating in a Wild West stage. AI is adding real value in some areas of marketing and erring in others.
The CMOs making real progress use a few common approaches:
Guardrails come before scale.
AI applications center on structured, measurable work, not in brand-defining roles.
A marketer's perspective is essential for narrative, creative, and strategic decisions.
Teams use AI as an accelerator, not as an autonomous operator.
AI can speed up analysis, widen creative exploration, and expand the work. Human judgment ensures decisions stay grounded.
If your team is determining where AI needs tighter guardrails, we can help you build a roadmap that’s practical and sound. Let’s map the opportunities together.