The Devil Wears Data: How AI is already shaping branding

Let's break down how AI powers marketing - and why human oversight still makes or breaks the experience.

Photo by yang miao on Unsplash

Do you remember the last Super Bowl?

Not the game (touted as a cinematic, symbolic blowout of good vs. evil) - the ads. There were exaggerated, disembodied facial features, like dancing mustaches and rogue eyebrows. In one, the musician SEAL was transformed into a literal half-man, half-seal.

Social media was quick to blame AI. Too many coincidences. Too many oddities. Surely these weren’t human-made.

As it turns out, this was a backlash. Here While AI was used behind the scenes, humans were still at the helm of the creative process. Still, it revealed something important: there’s a growing distrust of AI in branding, and a premium placed on human creativity. But AI is beginning to shape how brand stories are told - and how campaigns are built.

This post explores how AI is now integrated into the brand planning process - from photoshoot prep to storytelling to product prioritization - and how personalization and analytics amplify that. There’s real promise when it comes to efficiency and information flow. And, since these tools are still evolving, there still are watch-outs to keep an eye out for too.

Branding Execution: Behind the Scenes

If you’ve ever built a seasonal planning calendar or matrixed a pyramid of tactics to product launches, you know it’s equal parts storytelling and logistics. AI is beginning to streamline both.

For startups building their first brand, tools like LogoAI and Looka can generate a logo in minutes. On the creative planning front, Midjourney can now generate images as part of the concepting and moodboard process. Creative teams have found efficiencies with Wrike, a project management platform that uses AI to predict delays, assign tasks, and break goals into actionable steps. Persado offers AI-powered copywriting, analyzing emotional triggers to generate high-performing marketing language - especially for subject lines. Coherent Path uses AI to help retailers determine which content and product categories to feature, and when, based on customer behavior and revenue potential. Want a custom row of recommended products in an email? Certona tailors shopping experiences to different customer segments. And if you’re wondering how your brand is showing up on social, Sprinklr can monitor sentiment in real time.

Is it surprising that a brand might use some - or even all - of these tools behind the scenes? Or that there are even more that I did not list, including competitors for the above as well as others for a multitude of other functions (media, analytics, etc. - more on those later)? What may be more surprising is that most of them have been around since at least 2019, when I first began using them. These were early adoption days, with plenty of internal oversight. I’d argue that still matters. AI can improve processes, but we’re still responsible for the output.

What to Watch Out For in Branding Execution

Message Cohesion

A tool that optimizes subject lines might deliver a lift in open rates - perfect. But what if click-through rates drop?

These tools often don’t evaluate the full content of the email. Imagine a subject line that promises “60% off everything.” Intriguing, right? A lot of people would open that email. But if the body content showcases full-price new arrivals, that disconnect will lose the customer - and maybe some trust - and definitely a site visit.

Aesthetic Cohesion

Have you ever visited a website and seen a “We Think You’ll Like” row… only to be greeted with a mishmash of items that don’t visually belong together? Or maybe the row clashes with the balance of the homepage.

Most personalization tools can be configured with guardrails - for example, ensuring that only certain product images are grouped together or that visual styling aligns with a specific campaign or merchandising theme. But AI typically optimizes for a single KPI - often clicks - while brand stewards hope to balance multiple objectives, including storytelling, aesthetic consistency, and long-term equity. That’s why the brand team still needs to stay involved.

Missing Nuance

Tools to monitor brand sentiment in real time are powerful, but they’re not flawless. AI can flag a spike in mentions or emotional tone, but it might miss nuance. Sarcasm, irony, or playful banter often fly under the radar or get misinterpreted.

Imagine a flood of comments saying “Wow, amazing job 🙄” after a campaign launch. A human would see the sarcasm but a model may simply tag it as positive. And think of all the ways emoji combinations create a new language, especially with younger users. Here

Unless these patterns are constantly fed into an AI’s training set, proactively and not reactively, it’s tough to keep pace. That’s why AI-powered listening should always be paired with human context. Use the tool to surface signals, but don’t rely on it as your sole interpreter of mood.

Misaligned Objectives

Have you ever been in a meeting where one team says It worked! Look at these triple-digit lifts! and the rest of the meeting attendees say But it looks terrible or That’s not the product we’re proud of or Where’s my new launch product?

As a marketer, I recommend either being crystal clear if there’s a single KPI objective or, in the case of using AI, set shared KPIs and train AI platforms with full funnel goals in mind. AI tools may prioritize whatever will convert best (highly discounted or excess inventory), which could drown out a new product (with no history of selling) or a sustainability story (an awareness play). AI won’t account for your broader strategic goals unless you explicitly input them - and even then, ongoing oversight is crucial. And absent this feedback, any existing tensions between performance marketing and brand integrity might only be heightened by using AI.

AI Enhanced or Fake Reviews

What happens when the company - and the consumer - are both using AI? Reviews, especially for a new product, can greatly enhance adoption and sales. But now review platforms include a “Content Coach” which means consumers can get an assist in writing their review.

When reviews sound more like SEO copy than authentic feedback, trust erodes - even if the content came from a real customer. Or worse yet, what if it’s 100% fake, not independently written, and then sold to the company offering the product?

In August 2024, the FTC announced a new rule banning the creation and sale of fake reviews, including those generated by AI. This rule allows the FTC to impose fines up to $51,744 per violation. Here