OpenAI’s Copyright Trouble Just Got Much Worse

By Vicky Sidler | Published 3 February 2026 at 12:00 GMT+2

It started with a talking sponge in a meth lab.

No, really. That’s where OpenAI’s latest copyright nightmare begins. AI-generated versions of beloved cartoon characters have been popping up in questionable scenarios, and judges are starting to notice.

According to Futurism, the message is clear: the copyright issues AI companies have been dancing around are about to get serious. And no, “but it’s transformative” might not cut it anymore.

Here’s why that matters—even if you’re not in the business of cartoon meth labs.

TL;DR:

• Legal pressure on OpenAI is shifting from what it trained on to what it spits out

• Courts take visual content more seriously than text

• Characters like SpongeBob are huge commercial assets, and judges know it

• You might be liable too, depending how you use AI

👉 Need help getting your message right? Download the 5 Minute Marketing Fix.

Table of Contents:

Courts Have Been Tolerant... Until Now:

For years, AI companies have been stuffing models with everything they could find. Books, music, movies, random memes, your cousin’s blog post from 2009—all fair game.

So far, the legal defense has rested on transformative use. The argument? They weren’t stealing. They were learning. And that held up, more or less.

But visual content is another beast.

According to UC Berkeley copyright professor Pamela Samuelson, there’s a legal and emotional difference between text and images. A sentence from a novel might feel generic. A recognisable cartoon character, on the other hand, is expressive, distinct, and tied to a serious amount of money.

Think $16 billion in SpongeBob merch. Not exactly pocket change.

Why Judges Care More About Images Than Text:

There’s something about a picture that lands differently in court.

Judges are often more protective over visual IP because it carries immediate recognisability. It’s one thing to scrape a blog post. It’s another to generate a near-identical Mickey Mouse or SpongeBob doing things the original creators never imagined—and never approved.

If a user asks AI to make a parody, satire might be defensible. But if it just churns out a perfect copy of a trademarked character, things get tricky. Especially when that output leads to commercial gain for the user or reputational damage for the rights holder.

In simple terms, AI might be transforming input data, but it’s starting to output lawsuits.

The Slippery Slope From “Training” to “Copying”:

Here’s the catch: AI models don’t store images or paragraphs like a library. They learn patterns and generate new outputs based on what they saw during training.

That sounds safe—until the new output looks suspiciously like the original.

This is the tension legal experts are now highlighting. Is it remix or is it replica? If it looks too close to the real thing, judges may decide it’s the latter.

And that’s where the risk jumps from theoretical to financial.

You Might Be on the Hook Too:

Here’s where small businesses need to pay attention.

Futurism refers to an old Supreme Court ruling from 1984 involving Betamax (yes, the VCR). It’s part of a long-standing legal principle: if you build a tool that enables copyright infringement, you might be liable unless you take meaningful steps to prevent it.

OpenAI and similar platforms are testing the limits of that principle. But so are their users.

If you generate something with a recognisable character or logo and use it in your business—a Facebook ad, a website banner, a product label—you could be infringing copyright even if you didn’t mean to.

Intent isn’t the issue. Commercial use is.

This literally happened to a client of mine who asked AI to generate a retro-themed plumber mascot for their industrial plumbing business. What they got was a cheerful little man in overalls, standing on bricks, with a mustache large enough to qualify as a tax deduction. He was holding a wrench. He was jumping. And for reasons no one could explain, there was a turtle.

Now, I’m not a copyright lawyer. But even I knew we were two pixels away from a cease and desist.

So Think Before You Generate:

As a StoryBrand Certified Guide and Duct Tape Marketing Consultant, I’m not here to scare you off AI. I use it every day.

But I treat it like a blender. You don’t throw in raw meat, lemon peels, and your car keys and expect a gourmet smoothie.

If you're asking AI to generate anything client-facing, take a minute to check for copyrighted content. Is that cartoon actually a cartoon? Is that stock photo truly royalty-free? Does that logo look oddly familiar?

Because when it comes to copyright, ignorance isn’t a defence.

The Bottom Line for Small Businesses:

If you take away anything from this article, let it be this:

• Don’t trust AI to give you only safe content

• Avoid generating visuals that mimic known characters or brands

• If you’re using AI for marketing, do a copyright check just like you’d check for spelling errors

• Understand that legal risk applies to both the creators and the users

If you're trying to grow your business without accidentally stepping on a legal landmine, start by making sure your message is clear, consistent, and copyright-safe.

And if that message still needs work, I’ve got a free tool called the 5-Minute Marketing fix, which is designed to help.

👉 Download it free here.

Related Articles:

1. Targeted by PicRights? AFP's Copyright Demands Explained

If OpenAI’s copyright mess has you thinking about legal liability, this article gives step-by-step advice for responding to real copyright enforcement threats against small businesses.

2. AI Copyright Lawsuit: What Small Businesses Must Know Now

This piece tackles the other side of the copyright issue: the training data behind AI tools. Read it to get the full picture of why AI and fair use are legally shaky ground.

3. AI Workslop Is Wasting Time and Ruining Team Trust

Legal risk is one form of AI slop. Productivity risk is another. This article explains how low-effort AI content hurts internal teams and external trust.

4. Why Replacing Copywriters With AI Will Destroy Your Brand

If legal exposure doesn’t scare you, brand damage should. This article explains why removing human judgment from your content strategy is a shortcut to regret.

5. AI Hallucinations in Court—Big Trouble for Legal Trust

From fake citations to fake characters, this article explores the shared root problem: AI confidently produces false information. The cost of blind trust? Legal and reputational damage.

Frequently Asked Questions About AI, Copyright, and Small Business Risk

1. Can I use AI-generated images in my business marketing?

It depends. If the image is clearly original and doesn’t copy or resemble a well-known character, brand, or copyrighted work, it’s generally safer. But if it looks like a famous cartoon, logo, or photo, you could be infringing — even if you didn’t ask for it directly.

2. What if I didn’t know the AI content was copyrighted?

Unfortunately, not knowing doesn’t protect you. Copyright law focuses on whether the work was used without permission, not whether you realised it was protected. If it’s commercial use, the risk is even higher.

3. Are small businesses actually being sued for using AI content?

Yes. While most headlines focus on big companies or creators, copyright enforcement firms are already targeting small businesses over images and content pulled from the internet — including those generated by AI.

4. Is it safer to use AI for text than for images?

Generally, yes. Courts are usually more protective of visual content because it’s more expressive and easier to recognise. But text can still be risky, especially if it copies distinctive phrasing or includes copyrighted song lyrics, book passages, or scripts.

5. How can I check if AI-generated content is safe to use?

Reverse image search is your first line of defence. You can also use tools like Google Lens or TinEye. For text, run it through plagiarism checkers or check for unique phrases. When in doubt, don’t publish it without review.

6. Is OpenAI responsible if I get sued over an AI image?

Not necessarily. Legal experts say users can be held liable too, depending on how the image is used. If it’s used for commercial purposes or shared publicly, the risk increases — and so does your responsibility.

7. Can I claim fair use if I use AI content that resembles a known character?

Fair use is limited and complicated. It often doesn’t apply if your use is commercial, harms the original creator, or isn’t significantly transformative. Judges rarely accept “I didn’t know” or “I used AI” as a defence.

8. Should I stop using AI for marketing altogether?

No, but you should use it with care. AI can help you brainstorm, rewrite, summarise, or draft ideas. Just don’t assume everything it creates is safe, legal, or final. Treat it as a helpful assistant, not a copyright lawyer.

9. What’s the biggest risk with AI-generated visuals?

That they recreate something too close to a protected work. Famous characters, logos, or celebrity likenesses are heavily protected — and expensive to mess with.

10. How can I make sure my marketing is legally safe and effective?

Start by simplifying your message so you don’t need to rely on flashy AI content to stand out.Download the 5-Minute Marketing Fix and get one clear sentence that tells people what you do and why it matters.