Blog

<h3>A new generation of AI brand systems can now learn a brand well enough to change it. That doesn't mean that they should.</h3>A decade ago, I sat in the room with the leadership team from a prominent Silicon Valley organization. We were considering two finalist versions of their refreshed brand. The approaches had similar logos and color palettes with only stylistic differences in typography, so it should have just been an "apples or oranges" decision. But only one version was deemed correct as it felt like it captured the spirit of the organization. Most in the room could not articulate why, even though they agreed. It took the creative director explaining that the style of imagery in one approach "was too literal, not conveying experimental thinking" and that the better approach "had an unexpected blend of shapes and typography," better matching the spirit of the organization. Those nuances were key.That's what I keep coming back to as I watch a new category of "AI brand intelligence" platforms launch this year, each one promising to learn your brand well enough to keep it consistent, or in some cases, to evolve it, without a person in the loop for every decision. These systems are genuinely impressive. But some, in my opinion, are aiming at the wrong target.The new marketplace: From asset librarian to brand co-pilotAdobe, Jasper, Frontify, and Bynder have all shipped some version of this in the last several months, and it's worth understanding how differently they've approached the same problem. I've built a version of this too, <a data-mce-href="https://ideaovation.ai/ai-brand-systems" href="https://ideaovation.ai/ai-brand-systems" target="_blank" rel="noopener">Brand AI</a> by Idea Ovation.<a data-mce-href="https://business.adobe.com/products/brand-intelligence.html?sdid=GZKZXL82&amp;mv=search&amp;s_cid=701WQ00001jjmt7YAA&amp;s_iid=701WQ00001ZiGcNYAV&amp;edtamo=true&amp;s_kwcid=AL!3085!3!806203585826!e!!g!!adobe%20brand%20intelligence!23766311231!192596397261&amp;ef_id=Cj0KCQjw4JbTBhCoARIsALWUaBvTK9ut70sVZNvmZbNdD5p0I2otYt-qHYAQr-ITe8ex_srbj_NCJQ4aAmXkEALw_wcB:G:s&amp;gad_source=1&amp;gad_campaignid=23766311231&amp;gbraid=0AAAAADOODJT8PWKtZUMw0FgHPNjmLgahT&amp;gclid=Cj0KCQjw4JbTBhCoARIsALWUaBvTK9ut70sVZNvmZbNdD5p0I2otYt-qHYAQr-ITe8ex_srbj_NCJQ4aAmXkEALw_wcB" href="https://business.adobe.com/products/brand-intelligence.html?sdid=GZKZXL82&amp;mv=search&amp;s_cid=701WQ00001jjmt7YAA&amp;s_iid=701WQ00001ZiGcNYAV&amp;edtamo=true&amp;s_kwcid=AL!3085!3!806203585826!e!!g!!adobe%20brand%20intelligence!23766311231!192596397261&amp;ef_id=Cj0KCQjw4JbTBhCoARIsALWUaBvTK9ut70sVZNvmZbNdD5p0I2otYt-qHYAQr-ITe8ex_srbj_NCJQ4aAmXkEALw_wcB:G:s&amp;gad_source=1&amp;gad_campaignid=23766311231&amp;gbraid=0AAAAADOODJT8PWKtZUMw0FgHPNjmLgahT&amp;gclid=Cj0KCQjw4JbTBhCoARIsALWUaBvTK9ut70sVZNvmZbNdD5p0I2otYt-qHYAQr-ITe8ex_srbj_NCJQ4aAmXkEALw_wcB" target="_blank" rel="noopener">Adobe Brand Intelligence</a> builds a structured "brand ontology" from a brand's guidelines and assets, plus the approvals and rejections that capture its tribal knowledge, then uses that model to instruct automated content assembly and even predict engagement before something launches.<a data-mce-href="https://www.jasper.ai/brand-iq" href="https://www.jasper.ai/brand-iq" target="_blank" rel="noopener">Jasper's Brand IQ</a> takes a lighter touch, learning a style profile from existing content and flagging off-brand language in real time, but the output still passes through a human editor before it publishes. <a data-mce-href="https://www.bynder.com/en/" href="https://www.bynder.com/en/" data-mce-selected="inline-boundary" target="_blank" rel="noopener">Bynder﻿</a> has mostly used AI to make a traditional asset library smarter and faster to search, useful, but not a system making creative judgment calls.<a data-mce-href="https://www.frontify.com/en" href="https://www.frontify.com/en" target="_blank" rel="noopener">Frontify</a> has taken the position I find most interesting, essentially arguing against the category's own premise. Its "Human-Agent Collaboration Framework" holds that the interface for humans should remain human, and that AI's job is to make the brand's knowledge retrievable, not to make the calls itself.<a data-mce-href="https://ideaovation.ai/ai-brand-systems" href="https://ideaovation.ai/ai-brand-systems" target="_blank" rel="noopener">Brand AI</a> is my own entry into this category, and I mention it here for contrast as much as anything else: where the four platforms above are enterprise-scale systems built for organizations that already have a brand team in place, Brand AI is deliberately leaner, cheaper, and more task-focused, built for smaller organizations, smaller systems, and a narrower set of jobs, without trying to be everything Adobe's stack is trying to be.Where I disagree: When AI starts driving instead of informingHere's my candid take on this. Somewhere in that lineup, at least one vendor has crossed from "AI helps enforce the brand" into "AI directs where the brand goes next." One company's own language, automated assembly, brand enforcement running without a person at that step, engagement prediction shaping what gets made before it's made, describes a system where the AI isn't just checking your work anymore. It's increasingly doing the work of deciding what the brand should say and how it should evolve, with humans supplying a training signal rather than a judgment.The appeal makes sense. Enterprise marketing teams are increasingly downsized and drowning in content demand. Anything that promises to keep a thousand pieces of content on-brand without a thousand human reviews looks like relief! But I'd argue against letting any AI system, however sophisticated its ontology, hold the pen on where a brand's identity moves next. Not because the technology is bad. Because the thing it's being asked to steward is bigger than what an ontology, however well-built, can currently capture.A brand is not its assets. It's a set of nuanced sensibilities.Nearly thirty years into this work, across brands as different as Apple, Coca-Cola, and large university and public-sector institutions, the thing I've learned to trust least is a style guide's ability to explain a brand alone. A logo, a color system, a type family, those are the brand's clothes, not its personality. The actual brand lives in tonal choices that shift by half a degree depending on the moment, in the specific kind of humor a brand allows itself and the kind it doesn't, in knowing which technically-on-brief idea would still feel like a betrayal if you shipped it. That's sensibility, and sensibility is emotive before it's structural.AI brand ontologies are, by design, an attempt to encode exactly that, and Adobe calls it capturing "tribal knowledge" through implicit signals like approvals and rejections. I take that ambition seriously; it's a real advance over a static PDF nobody reads. But learning a pattern from a thousand past approvals isn't the same thing as understanding why the brand made those choices, and the difference shows up precisely at the moments that matter most: a cultural inflection point, a category-defining pivot, a joke that's a half-inch from landing wrong. Those are the moments a brand's evolution actually gets decided, and they're exactly the moments where I want a human's hand on the wheel, not a model's best statistical guess.Where AI earns its seat at the tableNone of this is an argument for keeping AI out of brand work. Quite the opposite, I think AI has a real and growing role, especially in the specific, bounded moments a brand needs to flex without needing to reinvent itself. A regional campaign that needs a slightly warmer tone. A product launch that needs the brand voice translated into a new channel's format. A rapid set of variations for a retail partner's ad units. These are small pivots and communication-specific adjustments, not identity decisions, and AI can handle a meaningful share of that work well today, with the right oversight.The line I'd draw is this: AI can accelerate and augment the expression of a brand that humans have already defined and are actively stewarding. What it shouldn't do, at least not yet, is originate or approve the evolution of that brand's identity on its own.Why I built Brand AI differentlyThat distinction is exactly why I built <a data-mce-href="https://ideaovation.ai/ai-brand-systems" href="https://ideaovation.ai/ai-brand-systems">Brand AI</a> the way I did. Most of the platforms above are enterprise-scale tools, priced and architected for organizations that already have a creative director, a brand team, and a governance function, adding AI on top of human infrastructure that already exists. The businesses I work with most often don't have that infrastructure. A growing company with one product and no creative director isn't well served by a system built to manage nuance at Adobe's or Coca-Cola's scale, and it's just as poorly served by a system that lets AI make brand decisions nobody senior is checking.Brand AI is built to sit in that gap. It's a virtual creative director, production designer, and brand manager that augments and accelerates human creative work rather than replacing it, offering direction, strategic point of view, and consistency oversight for teams that don't have a creative leader in the room, without taking over the production or the final call. It scales down to a single product with no formal brand system yet, and scales up to a full asset library and template system with ongoing virtual oversight, but the strategic ideas and stylistic direction it offers are meant to be a collaborator's input, not an autonomous decision. Where some AI brand systems are building toward AI that assembles and predicts on its own, I've deliberately built toward AI that makes a small team's human judgment go further.What changes now, and what shouldn'tI'll fully admit the ground is moving fast. It's plausible that a system with a deep enough ontology of a brand's full history could someday evolve a brand the way an experienced human steward does. I'm not ready to say that door is permanently closed.But we're not there yet. The sensibility a strong brand carries, its emotive register and personality, still lives in human perception in a way no ontology has learned to hold. Until that changes, AI should be the sharpest co-pilot a lean team has ever had, not the one holding the wheel. Idea Ovation's <a data-mce-href="https://ideaovation.ai/ai-brand-systems" href="https://ideaovation.ai/ai-brand-systems">Brand AI</a> was built in partnership with <a data-mce-href="https://synaxis.ai/" href="https://synaxis.ai/">Synaxis</a>.<a data-mce-href="https://ideaovation.ai/blog" href="https://ideaovation.ai/blog">more i/o posts</a> &nbsp;| &nbsp;<a data-mce-href="https://www.linkedin.com/in/jjonesdirects/" href="https://www.linkedin.com/in/jjonesdirects/">about the author</a>

Who gets to evolve your brand? The case for keeping humans at the wheel

<h3>It's time to begin the transformation of your marketing with AI. But moving too fast isn't necessarily bold, it can be a liability.</h3>There are already moments in large-scale AI marketing rollouts where someone in a leadership is asking, "Why aren't we moving faster?" A fair question, on the surface. The tools exist and the vendors are eager. The board wants results. But the question that actually determines whether your AI transformation will succeed or become an expensive lesson isn't about speed. It's about timing.Timing is the variable almost no one discusses honestly in AI adoption conversations. We talk about tools, budgets, headcount, and ROI projections. But the sequencing of how you introduce AI workflows alongside existing marketing operations is what separates a successful transformation from a chaotic one. Get it right, you build confidence, institutional knowledge, and measurable efficiency. Get it wrong, and you get confused teams, degraded output quality, and an executive team that loses faith in the marketing leaders trying to deploy AI effectively.This is playing out across organizations of all sizes right now. The companies navigating AI marketing transitions well aren't necessarily the ones with the biggest budgets or the most sophisticated tools. They're the ones that understand the value of running new systems in parallel with legacy operations long enough to validate outputs, build trust, and make the transition feel earned rather than forced. That discipline and the patience to do it right is harder than it sounds.The parallel runway principleThe single most important structural decision in AI marketing transitions is committing to a genuine parallel runway period. This means your new AI-assisted workflows run simultaneously with your existing team and processes, not as a replacement, but as a validation layer.Here's what that looks like in practice. Your legacy content team continues producing at their normal cadence. Meanwhile, your AI-integrated workflow team or lab produces the same deliverables...same briefs, same formats, same intended outputs. Then you compare them. Not to embarrass anyone, and not to declare a winner, but to understand where the AI workflow is genuinely strong, where it needs human intervention, and where it simply isn't ready.This comparison period does several things at once. It gives you real performance data instead of vendor promises. It gives your legacy team members visibility into what's coming, which is both fair and strategically smart. People who understand the transition are far more likely to adapt than people who feel ambushed by it. And it surfaces the gaps that no demo or pilot program ever reveals: the brand voice nuances, the approval workflow friction points, the edge cases that only appear when you're producing at actual volume.The length of this parallel runway matters. Too short, and you're making decisions based on incomplete data. Too long, and you're carrying double the operational cost without moving forward. For most mid-to-large marketing organizations, a meaningful parallel period runs somewhere between 6 months to a full year, depending on the complexity of your output mix and the maturity and curation of your AI toolset. Not as a universal rule, but a reasonable starting point. And based on the level of AI change that is our reality, a permanent experimentation effort should continue to keep your team aligned with the latest developments and opportunities. This could be an internal team or an outside consultant.Sequencing by output type, not by departmentOne of the most common mistakes I see is organizations structuring their AI rollout around departmental boundaries rather than output types. They say, "We'll start with the social team, then move to email, then tackle long-form content." That logic feels administratively tidy, but it creates uneven results and makes it harder to build cross-functional confidence in the new system.&nbsp;A more effective approach is to sequence by the complexity and risk profile of the output itself. Start with deliverables that carry the lowest brand risk and the highest volume: performance ad copy variations, metadata, SEO-driven content scaffolding, templated campaign briefs. These are areas where AI genuinely excels and where a subpar output doesn't carry catastrophic consequences. Success here builds organizational confidence quickly.Then move to medium-complexity outputs: campaign concepting support, content calendars, audience segmentation briefs, first-draft long-form content that gets significant human editing. This is where your team starts developing real intuition about how to work with AI effectively: what prompts work, what review steps are non-negotiable, and where the human creative layer adds the most value.High-complexity, high-brand-sensitivity outputs like flagship campaign creative, executive thought leadership, brand storytelling, customer-facing communications that carry significant emotional weight, these come last, and honestly, they may never be fully AI-driven. That's not a failure. That's an honest assessment of where the current technology actually is. The human creative product, the ability to read cultural nuance and genuine emotional resonance, remains the part of this work that AI tools still struggle to replicate. Acknowledging that openly is a strength, not a concession.This sequencing approach also makes it easier to re-skill team members progressively. Instead of asking everyone to change everything at once, you're introducing new workflows in layers, which is both more humane and more effective from a change management standpoint.The trust-building problem nobody budgets forHere's something that rarely appears in AI implementation budgets: the cost of trust-building. Not just with your team, but with your customers, your internal stakeholders, and your own leadership.Your marketing team needs to trust that the AI transition isn't just a prelude to mass layoffs. Your customers need to trust that the quality and authenticity of what you're producing hasn't been quietly hollowed out. Your CMO and CFO need to trust that the investment is producing measurable results, not just activity.Each of these trust relationships requires intentional work, and that work takes time. With your team, it means being transparent about the roadmap. Not just the tools you're adopting, but the roles you're evolving, the skills you're investing in, and the honest timeline for how the team structure will change. People can handle difficult news when they feel respected and informed. And if you offer to re-skill the legacy team with the latest AI tools, even those not retained will appreciate being trained and ready for the market. For many, this latest turn in the road will redirect them towards another career path or retirement.With customers, the communication question is increasingly important and genuinely complex. There's a real difference between AI-assisted work and AI-generated work, and that distinction matters to a growing segment of your audience. Being clear about how you're using AI, not defensive, not performatively humble, just honest, is becoming a brand differentiator. Companies that figure out how to communicate this well will have a real advantage over those that either overclaim their AI capabilities or quietly hope nobody asks.With internal stakeholders, the parallel runway data is your best asset. When you can show a CMO or a CFO side-by-side output comparisons, production timelines, and cost-per-asset metrics from your parallel period, you're not asking them to believe in AI on faith. You're showing them evidence. That's a fundamentally different conversation, and it's one that tends to unlock both patience and investment in ways that enthusiasm alone never does.Agentic systems are nearly here, and the timing question gets more complexThe current moment in AI marketing tools is interesting because we're in a transitional phase. Most organizations are still working with point solutions like an AI writing tool here, an image generation platform there, maybe a predictive analytics layer on top of their existing stack. That's manageable. Complex, but manageable.What's coming and what's already beginning to arrive from platforms like Adobe and Salesforce is a different order of magnitude. Agentic AI marketing systems that can autonomously manage multi-channel campaign execution, content personalization at scale, and real-time performance optimization are moving from concept to product. These aren't incremental improvements to existing tools. They're a fundamentally different operational model, and they will require organizations to have already built the internal competency to evaluate, validate, and govern AI outputs at speed.The organizations that will navigate that shift most successfully are the ones building their transition discipline now. Not because today's tools are the same as tomorrow's, but because the organizational muscles you develop during this current phase, things like parallel validation, output-type sequencing, trust-building, re-skilling, are exactly the muscles you'll need when the more powerful systems arrive. The companies treating this period as a learning investment will be positioned to move quickly when agentic platforms mature. The ones who skipped the discipline and just swapped in point solutions without building real transition competency will find themselves starting over, except now the stakes are higher and the systems are harder to govern.The patience that actually accelerates outcomesThe counterintuitive truth about AI marketing transformation is that the organizations moving most thoughtfully, not most aggressively, are the ones achieving durable results fastest.Speed without structure creates false starts. A team that adopts AI tools under pressure, without a real parallel validation period, without clear output sequencing, and without honest communication to staff and customers, will likely hit a wall. The outputs won't meet brand standards. The team will lose confidence. Leadership will question the investment and the marketing leadership. And then you're not just behind, you're behind and rebuilding trust at the same time, which is a significantly harder problem than the one you started with.The organizations getting this right are treating the transition as a genuine operational build, not a software installation. They're investing in the parallel runway and sequencing thoughtfully. They're having the hard conversations with their teams early. They're building the institutional knowledge that makes AI genuinely useful rather than just technically present.Timing isn't about moving slowly. It's about moving in the right order. And right now, for most large marketing organizations, the right order starts with planning and discipline, runs through validation, and earns its way to scale. <a data-mce-href="https://ideaovation.ai/blog" href="https://ideaovation.ai/blog">more i/o posts</a> &nbsp;| &nbsp;<a data-mce-href="https://www.linkedin.com/in/jjonesdirects/" href="https://www.linkedin.com/in/jjonesdirects/">about the author</a>

Timing is everything when onboarding AI into your marketing operations

<h3>The public is getting better at noticing synthetic AI visuals and audio. Is it time to hit pause on these types of AI-generated assets, or are we just doing it wrong?</h3>Last week I was scrolling through my YouTube feed and stopped dead at an ad for a snack product. The spokesperson looked like they'd been sculpted from matte plastic. The skin had that tell-tale AI smoothness. It was too perfect, too even, completely devoid of the micro-texture that makes a human face feel real. The hands and eyes were subtly wrong. I stared at it longer than I would have stared at a good ad, but not for the right reasons.It raised two questions that I haven't been able to shake. First: was this a budget problem, a skills problem, or is some level of obvious AI "fakeness" quietly becoming an accepted aesthetic, in the way that Pixar's stylized sheen became the defining look of animated cinema for a generation? And second: has the industry's rush to deploy AI-generated imagery outpaced the technology's ability to actually deliver on the implicit promise of photo-realism?I think we're at an inflection point. And the industry is not handling it gracefully.The Uncanny Valley Has a PR ProblemThere's a well-documented psychological phenomenon called the "uncanny valley," that deeply uncomfortable feeling you get when something looks almost human, but not quite. Robotics researchers identified it decades ago, but it has become a defining challenge of AI-generated marketing in 2025 and 2026.South Korea's GS25 convenience store released an AI-generated campaign that drew <a data-mce-href="https://www.campaignasia.com/article/ai-ads-and-the-uncanny-valley-problem/10e51cymjf3wpvkkmxu4qb5gao" href="https://www.campaignasia.com/article/ai-ads-and-the-uncanny-valley-problem/10e51cymjf3wpvkkmxu4qb5gao" target="_blank" rel="noopener">immediate public backlash</a>; "creepy," "soulless," and "plastic" were among the reactions, and it became a case study in what happens when a brand outsources emotional nuance to a generator. Coca-Cola, with all its resources, has now stumbled twice. The comparison between its 2024 and 2025 holiday campaigns reads like a strategy shift from technical ambition to aesthetic course correction, driven by the need to resolve the very same uncanny valley problem. Yet critics still found the end product "<a data-mce-href="https://hyperight.com/holiday-hype-or-digital-slop-coca-colas-2025-ai-ad-sparks-industry-split/" href="https://hyperight.com/holiday-hype-or-digital-slop-coca-colas-2025-ai-ad-sparks-industry-split/" target="_blank" rel="noopener">crummy looking</a>" despite a year of significant advances in AI video technology.&nbsp;This is not a small-brand problem. This is an industry-wide reckoning with the gap between what AI image and video tools promise and what they actually deliver when the ask is photorealistic human beings.Are Consumers Getting Smarter, or Are We Getting Sloppier?Here's where the data gets genuinely interesting, and a little contradictory.A <a data-mce-href="https://clutch.co/resources/ai-photos-brand-usage-consumer-trust" href="https://clutch.co/resources/ai-photos-brand-usage-consumer-trust" target="_blank" rel="noopener">Clutch survey</a> found that 57% of consumers couldn't identify AI-generated photos when tested, even though 66% were confident they could spot them beforehand. After the test, their confidence dropped to 56%. On the surface, that might sound like good news for marketers using AI. But read it again: confidence dropped. People walked away from the test more suspicious than when they started. And suspicion, once it enters the frame, is corrosive.The same study found that 84% of consumers want brands to disclose AI imagery, and trust drops sharply when disclosure is missing. Meanwhile, <a data-mce-href="https://www.nim.org/en/publications/detail/transparency-without-trust" href="https://www.nim.org/en/publications/detail/transparency-without-trust" target="_blank" rel="noopener">research</a> from the Nuremberg Institute for Market Decisions found that ads described as AI-generated were perceived more negatively than identical ads presented as human-made, with participants showing less inclination to click or engage with the product.So even when consumers can't tell the difference, the mere knowledge that AI was involved changes their behavior. That's a strategic problem that no improvement in render quality is going to solve.The Style Question: Sloppy Marketing or a New Aesthetic?Back to my plastic snack spokesperson. I keep coming back to that first question: is there an emerging AI aesthetic that audiences might eventually accept the way they accept animation or CGI?It's not a crazy premise. Every visual medium develops its own signature look, and sometimes that look becomes the point rather than a limitation. Early CGI in film was laughable compared to live action, and then Toy Story made the artificiality part of the magic. Could there be a version of that for AI-generated advertising?There are already cases that suggest yes — but with a crucial caveat. <a data-mce-href="https://marketingagent.blog/2025/11/02/virtual-influencers-vs-human-influencers-a-data-driven-comparison-for-2025/" href="https://marketingagent.blog/2025/11/02/virtual-influencers-vs-human-influencers-a-data-driven-comparison-for-2025/" target="_blank" rel="noopener">Dior's collaboration</a> with virtual influencer Noonoouri, known for her animated, illustrative style rather than hyperrealism, reportedly exceeded human influencer benchmarks by 42% in campaign ROI while cutting content production costs by 76%. Marketing Agent Blog The key was that Noonoouri doesn't try to be real. She's a fashion cartoon, and she's transparent about it. Experts in the space now <a data-mce-href="https://marketingagent.blog/2025/10/29/from-calvin-klein-to-gucci-how-top-brands-are-using-ai-avatars-to-drive-revenue/" href="https://marketingagent.blog/2025/10/29/from-calvin-klein-to-gucci-how-top-brands-are-using-ai-avatars-to-drive-revenue/" target="_blank" rel="noopener">recommend stylized realism</a> over photorealistic replication specifically to avoid the uncanny valley effect.&nbsp;That's a meaningful distinction. The brands getting punished aren't those leaning into a clearly synthetic aesthetic, they're instead the ones attempting photorealism and falling short. There's a massive difference between "this is obviously a stylized AI character" and "this is supposed to look like a real person but something is deeply wrong."The "AI Phobia" in the RoomI want to name something that doesn't get talked about enough: the irrational end of the anti-AI backlash. I call it AI phobia, and I see it regularly among senior marketing colleagues who have written off all AI-generated visual and audio content as categorically unacceptable. The fear is real, but it's often rooted in a misunderstanding of what the technology can and can't do, and in which contexts.Not all AI-generated visual work is a plastic spokesperson hawking snacks. There are AI-generated campaigns out there in artistic, illustrative, or abstract styles that are extraordinary, work that would be prohibitively expensive or technically impossible to achieve by traditional means. Notably, the ability to detect AI imagery diminishes considerably among consumers <a data-mce-href="https://insights.wiseup.pr/76-of-us-consumers-unable-to-spot-ai-generated-images-in-new-test/" href="https://insights.wiseup.pr/76-of-us-consumers-unable-to-spot-ai-generated-images-in-new-test/" target="_blank" rel="noopener">aged 40 and above</a> Insights, which means the generation most likely to be making marketing decisions is also the most likely to overestimate how much the average consumer can spot. That mismatch leads to overcorrection.The solution isn't to ban AI from your creative workflow. It's to be strategic about where you deploy it and honest about what it can currently deliver.So, Should We Hit Pause?Here's my honest take: no, but with some real caveats.We should absolutely pause on AI-generated photorealistic human beings in consumer-facing advertising until the technology genuinely closes the gap. The uncanny valley in AI advertising isn't a glitch, it's a signal that emotional intelligence is failing. When a brand outsources emotional nuance to a generator, the audience notices instantly. Using AI to produce a "real-looking" person who ends up looking like a mannequin is not a budget saving, instead a <a data-mce-href="https://www.nngroup.com/articles/ai-ad/" href="https://www.nngroup.com/articles/ai-ad/" target="_blank" rel="noopener">trust cost</a>. And trust, once lost, is expensive to rebuild.But that's a very specific use case. For concepting, storyboarding, and internal comping, AI is a gift. For stylized, illustrative, or clearly non-photorealistic work, it can be exceptional. For abstract, environmental, or product-focused visuals that don't involve people, the quality bar is already high enough that the risk is much lower. The rule of thumb I'm applying in 2026: if the AI asset needs to fool the eye, it's likely not correct for the final media (use careful judgement). If it's making no such claim, proceed with strategy and disclosure.The Covenant Still AppliesI've written before about the need for an AI covenant with customers. This is a published framework for how and when you'll use AI, and what you'll disclose. That thinking applies here more urgently than ever. When AI is paired with strong human insight, cultural clarity, and transparent intent, it stops feeling uncanny and <a data-mce-href="https://www.campaignasia.com/article/ai-ads-and-the-uncanny-valley-problem/10e51cymjf3wpvkkmxu4qb5gao" href="https://www.campaignasia.com/article/ai-ads-and-the-uncanny-valley-problem/10e51cymjf3wpvkkmxu4qb5gao" target="_blank" rel="noopener">starts feeling inevitable</a>. That's the creative moment the best marketers are operating in right now.The brands getting hurt aren't the ones using AI. They're the ones using it carelessly to chase the cost savings without asking whether the output actually serves the audience.The technology will get better. It always does. But in the meantime, the smartest move is to match the tool to the task, stop asking AI to fake things it can't convincingly fake, and be honest with your audience about what they're looking at. They're watching more closely than you think. <a data-mce-href="https://ideaovation.ai/blog" href="https://ideaovation.ai/blog">more i/o posts</a>&nbsp; |&nbsp; <a data-mce-href="https://www.linkedin.com/in/jjonesdirects/" href="https://www.linkedin.com/in/jjonesdirects/" target="_blank" rel="noopener">about the author﻿</a>

When AI looks like AI: The marketing case for knowing when to say no

<h3>It feels like the same movie we've seen before. But there are some new perils and big wins that lie ahead.</h3>In 2017, every enterprise software vendor had a blockchain strategy. Distributed ledgers would revolutionize supply chains, transform loyalty programs, and democratize data ownership. The promise was seductive: autonomous, trustless systems that eliminated intermediaries and reduced headcount. Management loved the cost-reduction narrative. Then reality arrived. Most organizations lacked the infrastructure to deploy blockchain meaningfully. The technology worked in narrow use cases—cryptocurrency transactions, certain supply chain applications—but the broader revolution never materialized because the gap between capability and operational readiness was massive.Now, we could be watching the sequel. Adobe and other platforms are pushing agentic AI as marketing's next revolution, and their 2026 Digital Trends report shows 80% of organizations believing breakthrough experiences will be "highly personalized and anticipatory" within a few years. Management sees the same thing they saw with blockchain: headcount reduction numbers that look good on a spreadsheet. But what they're trading for those savings—and whether their organizations are actually ready to make that trade—is a conversation most leaders are avoiding.The technology is becoming ready before most organizations are. That gap is where expensive mistakes happen, and where the human costs that don't appear in the initial analysis start to compound.The foundation problem no one's solvingBefore we get to autonomous agents replacing parts of your marketing team, let's talk about what needs to be in place first. Not as bureaucratic checkboxes, but as the foundation that separates outcomes that move the needle from ones that embarrass you publicly.Adobe's research confirms what I see in practice: despite widespread pilots with generative AI, only a minority of organizations have actually integrated it across multiple functions, and even fewer have embedded it enterprise-wide. The gap between toolset experimentation and deployment isn't a technology problem. It's an operational maturity problem that most teams haven't acknowledged yet.Here's what actually matters. Governance that has teeth—someone who owns what the agent can publish, who gets called when it goes sideways. You need to define accountability before you need it, not after. Brand intelligence that goes beyond static guidelines—voice, tone, messaging hierarchy, the things you'd never say. These need to be encoded, testable, and enforced. Your AI agent can't read a PDF and absorb your brand soul.Clean data with real consent. Agentic marketing runs on customer data, and before your agent touches a single contact, you need a unified CRM, clean IDs, and GDPR and CCPA compliance locked in. A poorly scoped agent that contacts opted-out customers isn't a technical glitch—it's a liability that makes headlines. According to Adobe's research, only 44% of organizations say their data quality is adequate for AI, and 52% admit that poor data unification limits AI advancement. That's not a foundation, that's quicksand.Channel access needs to be intentional, not open. Scoped API keys, spend caps, approval thresholds. Your agent doesn't need admin rights to everything. The ability to autonomously launch a major campaign without human sign-off is a risk, not a feature. Compliance needs to be baked in from day one—disclosure rules, FTC and ASA requirements, IP clearance for images and copy.You need baselines before deployment. If you don't know where you started, you can't prove where AI took you. Establish performance benchmarks first: conversion rates, cost-per-acquisition, brand sentiment. Then measure what changes. And you need a kill switch with a tested rollback procedure for each channel and a communication plan for when something goes wrong.Most critically, the humans working alongside the agent need to understand what it can and can't do well enough to catch errors, escalate issues, and step in when the output drifts. AI literacy in your marketing team isn't a nice-to-have. It's the last line of defense.Get these foundations right, and agentic AI becomes a genuine force multiplier. Skip them, and you're just generating risk at scale.Where limited autonomy actually works today I haven't seen an autonomous system that impressed me yet. But some are getting closer. Adobe's case studies with their Journey Orchestrator product are approaching meaningful autonomy in narrow domains.The successful applications I've seen share specific characteristics. They operate in constrained domains with clear success metrics, reversible decisions, and human oversight built in. Agentic systems can already replace repetitive, pre-formatted or templated asset production—visuals, copy, videos. I'd commit to the highest repetition tasks first, like placing product shots into pre-formatted online ads that retailers run constantly. You still need human QA, but the speed for these tasks is incredible and the quality is already good enough.Bid management and media optimization with clear KPI boundaries works. Customer service routing and initial response in high-volume, low-complexity scenarios works. Content variation generation within strict brand parameters works. What these share is that they're narrow, rule-based applications where the cost of error is low and the ability to course-correct is high.What doesn't work yet is the broader promise—autonomous agents making strategic decisions, creating breakthrough creative work, or managing complex customer relationships end-to-end. As Dan Gardner, co-founder of Code and Theory, argues, "AI is now the baseline infrastructure for knowledge work. Writing, research, prototyping, summarizing, coding. That's individual leverage. It makes people faster. The more profound shift is collective leverage." The real opportunity, he suggests, is in collaborative intelligence where AI elevates realignment rather than replacing it.The headcount math that looks better on spreadsheets than in reality There's a financial argument for agentic AI. The potential to replace two to four FTEs in specific workflow areas. Immediate budget relief. 24/7 operation. When a CFO comes to you with that math, you need to have a different conversation than the one they're expecting.Before organizations lay off 50-60% of their marketing team, they need to understand that these systems aren't perfected yet—they're not even good enough yet. What makes more sense is hiring a shadow team to start ramping up, tuning, and teaching the agentic systems your brand, quality expectations, and guardrails. As performance starts hitting the mark, ramp that up while ramping down the legacy team.The spreadsheet doesn't capture institutional knowledge loss, creative judgment, relationship equity, or quality degradation over time. It definitely doesn't capture what happens to the team that remains. Morale suffers. Trust erodes. Risk-taking disappears. And here's a question most leaders aren't asking: if senior marketers are running systems and juniors aren't needed, how do we train the senior leaders for the future?Adobe's research found that one in four customers now turn to AI-powered platforms as their primary source for information and purchase decisions—surpassing brand websites and reviews. When AI is increasingly competing against AI in the discovery phase, do you really want your marketing output to be entirely AI-generated? Gardner points out that "delegation is overtaking browsing as the key vehicle for discovery." People ask LLMs questions during discovery, validate socially where trust forms, and transact wherever it makes sense. In that environment, what advantage does human creativity provide?The answer: the difference between generic and custom. Between something that works because an algorithm optimized it and something that works because a human understood context that algorithms miss.The catastrophic errors that are coming Organizations that over-rely on these new systems and drop headcount too fast are going to make hugely embarrassing and perhaps catastrophic errors. It will happen. And consumers are ready to punish organizations who make headlines for an over-reliance on AI or huge layoffs for the same reasons. It's the same reaction people get to an AI bot or weird AI images with too many fingers—there's an AI-phobia growing in the public and they will take action, especially against consumer brands.We're approaching a 1995 moment, like when internet companies fired vast numbers of marketing and brand leaders because they weren't embracing the internet. The difference is that the internet actually was ready for widespread deployment. Agentic AI isn't—not yet. The organizations that move too fast will become cautionary tales.What moving deliberately actually looks like If an organization came to me today wanting to deploy agentic AI in their marketing operations, here's the sequence. First, get your house in order: strong brand guidelines, well-organized assets and templates, clear marketing goals with metrics including staff savings goals, a library of approved materials for different platform applications. You need technical overviews of what's running your website and other systems to understand compatibility and efficiency opportunities.Then comes the overlap phase. I'd do a one-year ramp-up in parallel with the legacy team ideally, with no less than six months to do the same. Team reductions can happen in stages as certain milestones are reached. This isn't about being cautious for caution's sake. It's about building something that actually works instead of something that looks good in a quarterly earnings call.Adobe's research confirms that "AI capabilities are evolving faster than organizations can keep up, and customer expectations are shifting just as quickly." The constraint used to be technical feasibility. Now, as Gardner observes, "it is imagination." But imagination without operational foundation is just expensive daydreaming.The honest timeline is twelve to eighteen months of operational work before responsible agent deployment. That's not what management wants to hear, but it's reality. And the leaders who acknowledge that reality—who build foundations first, pilot narrowly, and honestly calculate both financial and human costs—are the ones who'll actually deliver on AI's promise rather than just its hype.Tread carefully, move deliberately The technology will be ready before most organizations are. That gap is where the expensive mistakes happen, where the human costs that don't appear in the initial cost-benefit analysis start to compound, and where the distance between what vendors promise and what operations can actually deliver becomes painfully obvious.The question isn't "can we deploy agents?" It's "should we, and at what cost?" The leaders who win won't be the ones who move fastest. They'll be the ones who move deliberately—who understand that the goal isn't autonomous marketing, it's augmented teams that deliver what AI alone cannot: custom quality at optimized cost.We've seen this movie before. The ending doesn't change just because the technology does. What changes is whether you're ready when the credits roll. <a data-mce-href="https://ideaovation.ai/blog" href="https://ideaovation.ai/blog">more i/o posts</a> &nbsp;| &nbsp;<a data-mce-href="../" href="../" target="_blank" rel="noopener">about the author</a>

Agentic AI in marketing: Is it this decade's overhyped moment?

<h3>Marketing has always been prone to spin and the distortion of reality. Now we're just debating whether it should come with a disclosure label.&nbsp;</h3>I've been in the marketing business since the 1990s, and I can tell you with absolute certainty: we've never been completely honest about how we create the images that sell products. Not in the way people think we have, anyway.Here's what fascinates me about the current panic over AI in marketing: 73% of consumers say they value authenticity in brand communications, yet 82% accept enhanced product photography as standard practice. That gap tells you everything you need to know about the authenticity debate we're having right now. We're arguing about the wrong thing.When I started at Macy's in retail advertising—before design work was even digital—models' clothes were pinned so they fit better than they ever would when purchased. After photography, both the clothing and the models were heavily airbrushed. What customers saw 40 years ago was already a synthetic reality. They may or may not have realized this, but they embraced it anyway. Those impossibly skinny models wearing clothes that fit in impossible ways? Everyone knew it was unrealistic, but we all went with it.So when people ask me if using AI to create marketing content is fundamentally dishonest, I have to push back on the premise. The question isn't whether AI is honest or dishonest. The question is: why are we suddenly demanding a level of transparency we've never required before? The Reality of Marketing Enhancement: A History We've Conveniently Forgotten Let me walk you through what we've already accepted as standard practice in marketing, because I think we've developed collective amnesia about how this industry actually works.Product Photography and RetouchingThe airbrushing I mentioned from my Macy's days was just the beginning. Today, every product image you see has been enhanced in ways that would shock most consumers. Colors are adjusted, textures are perfected, and imperfections are eliminated. We've been doing this for decades, and nobody calls it dishonest anymore. It's just... marketing.Automotive AdvertisingHere's something that should blow your mind: approximately 90% of automotive advertising today uses CGI, not actual photography. Those cars you see racing down impossibly scenic roads with impossibly shiny paint and perfectly aligned panels? They're 3D renders. The cars often don't even exist yet when the ads are created.Can a little Toyota really do 100 mph on a twisty mountain road like those renders show? Of course not. Are customers fooled? Maybe some are. But we all understand on some level that we're looking at an idealized version of reality. We've made peace with that enhancement because it serves a purpose—it shows us the aspiration, the possibility, the best version of what we might experience.Packaging and Product DesignThe same 3D rendering technology has been standard in packaging design for years. That cereal box, that shampoo bottle, that tech product packaging—it's all been rendered in ways that create an impossibly perfect version of the physical product. The colors are more vibrant, the surfaces more flawless, the proportions more ideal than what you'll actually hold in your hands.Fashion Industry StandardsBeyond the pinning and airbrushing I experienced early in my career, the fashion industry has been using 16-year-old models and presenting them as if they're in their twenties or older. Post-production techniques have been standard practice for so long that we don't even think about them anymore. Every fashion campaign you see has been enhanced, adjusted, and perfected in ways that have nothing to do with reality.The progression from airbrush to Photoshop to AI isn't a revolution—it's an evolution. Each tool made the enhancement process faster, more accessible, and more sophisticated. But the fundamental practice? That's remained constant.The Mango Moment: When AI Became Undeniable This brings me to a campaign that I continue to look at in amazement: Mango's AI-generated campaign for their teen line.Mango, a global fashion retailer, created an entire campaign using generative AI as part of their Strategic Plan for 2024-2026, which explicitly aims to create value through technological advancement. The results were stunning—so good that I keep going back to study them. It's possible there was some human retouching, just like we did with photos before, to perfect certain elements. But the core creation was AI-generated.Here's what's remarkable: there wasn't the backlash that hit American brands like J.Crew and Guess when they tried similar approaches.J.Crew has a loyal customer base, and those customers felt betrayed by what they perceived as a lack of transparency and authenticity. (There's that word again—authenticity.) But I suspect much of the anger came from people in the design and photography industries who saw their jobs threatened.Guess faced similar criticism when they featured AI-generated models in Vogue. A viral TikTok video exposed the "fake women," leading to public outcry about job displacement and unrealistic beauty standards. Critics were concerned about real models and photographers losing work.But Mango? They executed it in a way that worked. The campaign was part of a transparent strategic initiative. The quality was exceptional. And critically, they understood their audience and market positioning.What does this tell us? That AI in fashion and marketing isn't going away, no matter how many people complain. Large organizations are taking this seriously, and the results are becoming incredible. The genie is out of the bottle.Now, I'll be honest—I do question clothing products generated entirely in AI. Does it really look and fit that way? It can't be perfect. But then I remember the pinned clothes and airbrushing at Macy's, and I realize we've always had this gap between representation and reality.The "Human-Washing" We've Always PracticedThere's a term floating around now: "AI-washing"—brands hiding their use of AI. But here's what nobody wants to admit: we've always practiced "human-washing" in marketing.We've always obscured the production process. We've never disclosed every tool, technique, or enhancement we used. When was the last time you saw a fashion ad with a disclaimer: "Model's skin retouched, clothing pinned for better fit, colors enhanced, lighting artificially perfected"? Never. Because that's understood to be part of the creative process.The dishonesty, if you want to call it that, has always been there. People have a phobia of AI right now, and they're looking for disclosure. Companies know this, and they'd rather not admit they're using AI because it's bad PR. But it's happening anyway, which is why I argue for what I call an "AI covenant" with customers—a clear, published framework for transparency.The challenge is that it's going to take time for the public to understand what AI actually is. They might never fully understand, but they sure as heck know they don't like it right now. That's the reality we're navigating.Where the Ethical Line Actually Exists So where do I draw the line? When does enhancement become deception?I have real concerns about AI's depictions of physical products you're going to purchase. But I have the same concerns about 3D rendered items in general. It's potentially deceptive—it's not the real thing. We can only hope that creators do their best to accurately represent their actual products.The more profound ethical questions go beyond marketing. I was recently in Las Vegas watching a hologram of Whitney Houston performing her hits. I found it deeply unsettling. It's now possible to upload your skills, your thinking, your voice into AI—to create what many people call a digital copy. Who owns this? Who gives permission to make a digital copy of you?This is already happening. Little pieces of all of us are being integrated into AI, and there's nothing we can do about it right now. Who owns the rights to your likeness after death? We're not all celebrities with agents and lawyers. Who would own a digital copy of you after you're gone?These are the ethical questions that are going to be difficult to answer. But they're also distractions from the immediate strategic reality facing marketing leaders.The Strategic Mindset Shift Marketing Leaders Must Make If you're a marketing leader reading this and thinking about AI integration, here's what you need to understand: we're approaching a split-in-the-road moment similar to the internet revolution of the late 1990s and early 2000s.I watched a lot of people lose their jobs when we switched to internet-based strategy in marketing. I kept mine and suddenly found myself in charge. The people who adapted, who retooled, who understood that their knowledge and creativity were being boosted rather than replaced—they thrived. The ones who resisted? They took the exit ramp and never came back.We're coming up on that moment again. Everybody who's on board with AI keeps going. Everybody who's not takes an exit ramp, and that'll be it.Here's what that mindset shift looks like in practice:Stop Debating Philosophy, Start Evaluating ToolsThe authenticity debate is a distraction from strategic opportunity. Your competitors aren't sitting around having philosophical discussions about the nature of reality in marketing. They're implementing tools that let them work faster, more efficiently, and more creatively than ever before.Understand the Workflow ImplicationsYou need to understand what these tools actually are and what they mean holistically to your workflow and future. A lot of people aren't even there yet, and those are the people who are just going to disappear. You can't make strategic decisions about something you don't understand.Address the Communication ChallengeFor internal agencies or agencies working with consumer products companies, there needs to be a conversation about transparency with customers. These customers are sensitive right now, and it can be a PR mess to have a leak about AI-generated marketing assets.I've already dealt with clients who think, "Oh, the AI's doing it, so why are you charging so much?" This is a complicated communication challenge. AI helps you do better work faster, but that doesn't mean it should be cheaper. In fact, this may lead us back to value-based pricing, making hourly rates irrelevant since the time it takes to do work is becoming less meaningful.Create an AI CovenantBefore you can communicate externally, you need an internal strategy. What will you use AI for? What won't you use it for? How will you monitor quality and appropriateness? What will you disclose to customers?Then you need an external covenant—a published framework that promises certain transparency when something is AI-generated. Now, this gets complicated because we're starting to use AI for every bit of our workflow. At what point does disclosure become meaningless because AI is everywhere?That's the conversation we need to be having, not whether AI is "authentic" or not.The Competitive Reality Nobody Wants to Acknowledge Here's the hard truth: if brands don't adapt to AI tools, they're going to fall behind in ways they can't recover from.The efficiency gap is opening up right now. The creative possibilities your competitors will have access to are expanding daily. The budget implications of not adapting are significant. We're not talking about a gradual shift—we're talking about a cliff.I think we're coming up to a moment where there will be a burst of opportunity to set up AI tools and workflows in companies. That will generate work. But once all that is set up, I do think there will be a drop in the workforce around marketing, similar to the year 2000 surge in the web industry and then the subsequent crash.This is a dire situation for a lot of professionals. If they don't embrace these tools and the efficiencies they enable, they won't just struggle to get jobs—they'll become completely uncompetitive in the marketplace.What We Should Actually Be Talking About If I could get marketing leaders to change one thing about how they're thinking about AI and authenticity, it would be this: stop treating AI as a unique ethical challenge and start treating it as what it is—the latest evolution in a long history of creative enhancement.We've always modified images of our products. We've always used flowery copy and exaggerated the value proposition. We've always created idealized versions of reality. This is not something new.The real conversations we should be having are:How do we communicate with customers about our use of AI in ways that build trust rather than erode it? How do we help our teams adapt to these tools so they can do better work with less stress? What's our internal strategy for responsible AI use, and how do we monitor it? How do we navigate the sensitivity around replacing human models and creators with synthetic alternatives? The replacement of people with synthetic people—that's where the real discomfort lies. That's what makes AI feel different from Photoshop or CGI. And that's a legitimate concern that leadership needs to address thoughtfully.But here's what I know from being in the trenches of AI implementation: these tools are helping people do great work faster with less stress. Your knowledge, capabilities, and creativity are all boosted by AI. You can do things that were never possible before.The question isn't whether marketing was ever honest. The question is whether we're ready to be honest with ourselves about what marketing has always been—and what it's becoming.The authenticity paradox isn't about AI. It's about our collective willingness to acknowledge that we've always valued the enhanced version of reality over the literal one. AI is just making that truth harder to ignore.And maybe that's not such a bad thing. <a data-mce-href="https://ideaovation.ai/blog" href="https://ideaovation.ai/blog" data-mce-selected="inline-boundary">more i/o posts</a> &nbsp;| &nbsp;<a data-mce-href="https://www.linkedin.com/in/jjonesdirects/" href="https://www.linkedin.com/in/jjonesdirects/" target="_blank" rel="noopener">about the author</a>

i/o blog

Who gets to evolve your brand? The case for keeping human...

Timing is everything when onboarding AI into your marketi...

When AI looks like AI: The marketing case for knowing whe...

Agentic AI in marketing: Is it this decade's overhyped mo...

AI and human-washing: Was marketing ever completely hones...