Don't put a target on your back 🎯

Hello marketers. Welcome to AI Marketing School, where we dish out the latest and greatest in AI-powered marketing.

In this week’s issue:

• AI Marketing Update: OpenAI’s GPU-melting viral episode

• The Stack: Words AI never uses

• The Stack #2: Free AI model directory

Onwards!

AI MARKETING UPDATE

🎨 AI hype meets copyright reality

If you were online last week, you probably saw it: dreamy, pastel-toned portraits in the style of Spirited Away, Princess Mononoke, and My Neighbor Totoro flooding X, Reddit, and Instagram.

The internet had a new obsession, and it was powered by OpenAI.

All you had to do was tell ChatGPT’s new image generator to create something “in the style of Studio Ghibli,” and you’d get eerily good results.

Even Sam Altman, OpenAI’s CEO, joined in, switching his X profile photo to a Ghibli-style version of himself.

The vibe? Let the chaos play out. And it did: over a million users signed up in just a single hour, overloading the system so badly that Altman said their GPUs were “melting.”

Ghibli is blatantly copyright (well, at least developing images in this style affirms Ghibli work is in OpenAI’s data; while a ‘likeness’ technically can’t be copyrighted, those images are a manifestation of a copyright breach).

The fact that Altman didn’t just block prompts relating to that keyword tells you something, too.

While the outputs do not breach copyright (at least by the conventional legal definition), many still commented how they constitute copyright manipulation on a massive scale.

Not everyone agrees with that, though. Is this a career distilled into a prompt? Or is AI creativity truly blooming in the public consciousness?

Either way, now, OpenAI has hit the brakes. Users began to notice that prompts referencing “Ghibli” no longer worked the same. Some seem blocked entirely. Others returned generic or muddled results.

OpenAI later confirmed it was taking a “conservative approach” and refusing prompts that replicated the style of living artists.

Another OpenAI viral marketing stunt?

Now, there’s a marketing masterclass here that OpenAI and Altman have played out before.

It goes like this: let a model launch with looser controls, watch it go viral, then nerf it later in the name of legal caution or policy shifts.

When OpenAI did voice mode demos, the same happened — millions of subs, then the product underdelivered for a while (it’s pretty good now, in fairness, but is still inhibited vs those first demos).

Just because AI companies get away with it doesn't mean everyone can

Users might get away with Ghibli-inspired selfies. But if you're a brand using this sort of imagery for commercial gain, you're putting a target on your back.

Studios like Ghibli might ignore the AI companies – they’re too big to fail right now – but they could go after businesses deploying images that blatantly defraud artists.

When you cross the line from fun to commercial, the risk becomes yours, not the AI's.

We can pull up Disney as an example. Notorious for vehemently protecting their copyright, they went after Etsy sellers in a series of legal attacks.

You can only imagine the absolute bloodbath that would ensue if the Ghibli moment happened with Disney imagery (not that OpenAI would ever allow that).

Plus, according to the U.S. Copyright Office, prompts alone don’t qualify as authorship as of this month. So not only might your AI-generated campaign content infringe on someone else’s rights – you might not even own it yourself.

The takeaway here is twofold:

1) Just because AI normalizes blurring copyright doesn’t mean marketers and business owners can.

2) And you may not be able to claim ownership over any high-value images you create with AI (logos, etc).

THE STACK

Words AI never uses 🚫

In past newsletters, I’ve focused a lot on words to avoid if you don’t want your content to feel AI. But what about words and phrases AI never uses, which humans do

Here's a guide to words and expressions that can make your AI-edited content instantly more human:

1. Colorful superlatives

AI tends to stick with basic positive descriptors with predictable divergence (e.g. transformative, game-changing). They are not always bad, and sometimes they are the best fit, but they are probably best avoided when possible.

• 🤖 AI loves: Effective, great, beneficial, optimal, valuable, high-quality, transformative, game-changing

• 🧠 Humans use: Brilliant, mind-blowing, superb, wonderful, fantastic, superb, out of this world, outstanding, incredible, amazing, wonderful

2. Industry slang & jargon

AI knows the dictionary definitions but misses the living vocabulary professionals actually use.

• 🤖 AI loves: Cost-effective marketing strategy, positive customer feedback, increased engagement metrics 

• 🧠 Humans use: Getting more bang for your buck, customers are eating it up, moves the needle, crushing your KPIs, blowing up your metrics

3. Cultural touchpoints

AI is generic, while humans naturally reference shared cultural knowledge.

• 🤖 AI loves: Similar to popular science fiction concepts, reminiscent of classic literature, akin to famous historical events 

• 🧠 Humans use: It's giving Black Mirror vibes, has serious Hemingway energy to it, peak 90s MTV aesthetic, Red Bull marketing swagger (ok there might be better ones, but this is what I came up with). Also, trendy phrases like, gives me/you the ick, hit me/you right in the feels. 

4. Bold claims vs. sitting on the fence

This is the big one! AI constantly hedges with academic-sounding qualifiers. Make your writing more confident. 

• 🤖 AI loves: May potentially improve, can be particularly effective, generally considered to be, it is worth noting that, generally overusing ‘typically’ 

• 🧠 Humans use: This strategy is pure gold, absolutely smashes/beats expectations, absolutely rock-solid, can’t be faulted, a truly outstanding example, first-class, best-in-class

5. Alternatives to "just" and not just"

Many will know that AI loves these crutch phrases when explaining concepts.

• 🤖 AI loves: This isn't just about increasing sales, but also..., Marketing is not just creating content, but also… (yawn)

• 🧠 Humans vary their transitions: This extends beyond numbers alone, Marketing encompasses everything from...,+ generally avoiding ‘not just x but y’ 

6. Strong negative language

AI tends to soften criticism, while humans are more direct.

• 🤖 AI loves: Suboptimal outcomes, less effective approaches, challenging implementation

• 🧠 Humans use: Total waste of money, absolutely tanked, completely missed the mark, epic fail

When editing AI content, it’s important to go beyond merely stripping it from AI hallmarks. Inject human content into it, too. It’s not a procedural thing – it takes genuine thought to do, which shows in the final product. 

Your audience can sense the difference — and as the research from last week showed, they respond better to content with that authentic human touch.

THE STACK #2

HuggingFace Spaces

If you've spent any time exploring AI models, chances are you've heard of Hugging Face.

At its core, Hugging Face is a platform for hosting and collaborating on machine learning models. But over the last year, it's grown into a kind of operating system for the open AI ecosystem.

One of Hugging Face’s most useful (and underrated) features is Spaces — a growing library of free, browser-based AI apps built by developers, researchers, and startups around the world.

Think of it as a directory of live AI demos: no installations, no sign-ups required. Just click, run, and test. For marketers, this makes it an incredibly fast way to explore what’s possible — from generating images to building brand assets or testing ideas before they’re fully productised.

While a useful playground for indie devs, major players are active here too, such as ByteDance (from TikTok), Stability AI, Meta, and Black Forest Labs, spun out of Stability AI.

There are absolutely tons of tools on there. You’ll need a free account at least, otherwise most are severely limited. Here are a few highlights from the full directory here:

• AI Comic Factory – Generate full comics from a single prompt, complete with panels and dialogue

• Hyper FLUX 8steps - Excellent fast text-to-image model

• IllusionDiffusion – Create trippy, high-quality illusion art for unique visual campaigns

• Open Meme Studio – Turn prompts into memes or avatar-style graphics instantly

• InstantID – Personalised image generation while preserving identity (great for brand avatars.

• Kolors Virtual Try-On – Overlay clothing on user-submitted images for ecommerce and product demos

• InfiniteYou-FLUX (by ByteDance) – Recraft photos while preserving facial identity

• TRELLIS – Generate scalable 3D assets from 2D images

• Hi3DGen – High-fidelity geometry generation from 2D inputs

• DeepSite – Generate full apps with DeepSeek, a large model that builds from ideas

One last thing

Spaces is also a window into how the AI ecosystem is evolving. Tools here often appear before they hit mainstream platforms, and they’re far less restricted in what they can do.

That makes it a useful spot to watch for trends that affect AI development before fully regulated products hit the mainstream market.

CONSULTANT’S CORNER

🗓️ Recommended AI marketing events

Whether you’re building with AI, leading a marketing team, or just trying to stay sharp in a fast-moving space, events are still one of the best ways to plug in.

Great events do more than just deliver content — they give you context. You get to see what other teams are testing, what’s actually working, and where the market is headed.

They’re also one of the few places where you can meet like-minded people, talk shop with folks solving the same problems, and build a network that isn’t purely algorithmic!

Some top events to check out:

•  AI Agents Summit (Virtual) — September 18–19, 2025: Laser-focused on AI agents, copilots, and autonomous systems. If you're experimenting with automating parts of your marketing workflow or interested in agent-based design, this one’s worth attending.

• AI for Marketers Summit (Virtual) — November 13–14, 2025: Created specifically for marketing professionals. From prompt engineering to campaign automation, it’s a practical look at how AI is being used right now in real teams.

• Data + AI Summit (San Francisco) — June 9–12, 2025: Hosted by Databricks, there’s plenty of marketing relevance here — especially around LLM workflows, AI tooling, and future-proofing your stack.

• Ai4 2025 (Las Vegas) — August 11–13, 2025: A big-picture AI conference covering finance, healthcare, retail, marketing, and more.

Hope you enjoyed this week’s issue. If you missed the last newsletter, you can read it here.

If you found it useful, please recommend it to a friend or colleague.

Until next time. Happy marketing.

—The AI Marketer