Can I protect the work I've made with AI? It's a fair question, and what most people don't realise is that ownership in the AI space isn't guaranteed. Then again, is anything ever guaranteed in life?
In a piece I wrote last month, "Unpacking Digital Rights. Who Really Owns Your Digital Twin?" I looked at how digital identities are becoming commodities and who owns these digital versions of us.
Today, I want to zoom in on an equally important issue, which is AI and copyright. Who owns what? What can you protect? And what happens if you get it wrong?
When AI Creates, Who Gets the Credit?
Let's start with the legal basics. Copyright law, in most countries, is built on one fundamental idea, which is that protection goes to human creators.
If a piece of content is generated entirely by an AI, with minimal or no human involvement, there's a high chance it won't be protected by copyright law at all.
That means you might pour hours into crafting the perfect prompt or building a dataset, only to end up with something you legally can't claim as your own. It's already a risk playing out across industries.
3 Real-World Lawsuits That Should Be Wake-Up Calls
New York Times vs. OpenAI and Microsoft (Publishing Industry)
The New York Times (NYT) filed a lawsuit against Microsoft and OpenAI in 2023, alleging that ChatGPT was trained using their articles without their consent and can now provide rewordings or summaries that directly compete with the original content published by NYT.
It has the potential to establish a new standard and is among the most important copyright cases in recent memory. This case is moving through the courts in 2025.
Scarlett Johansson vs. AI Voice Clone (Entertainment Industry)
Another example is Scarlett Johansson, who filed legal claims against OpenAI when it released a voice called "Sky" similar to hers.
It prompted new questions around vocal likeness, consent, and ethical AI usage. Given the ease of cloning voices via numerous apps that emerge every few weeks, we are going to see many cases in the press.
Authors vs. Anthropic (Literary and Tech Industry)
A lawsuit was filed against Anthropic, alleging that the authors' copyrighted works were used without permission to train the Claude AI models.
The case highlighted growing concerns within the literary world about consent and the boundaries of generative AI. Anthropic won over copyright, but it's still going to trial over how it got the books.
Isn't this just for creatives?
If you're a founder using AI to brainstorm product names, a marketer generating copy, or a business leader leveraging AI for client reports, this affects you too.
Every piece of AI-assisted content you publish or monetise comes with risk. If a human does not clearly author it, or you can't verify the source of the data it was trained on, you could be opening the door to legal action.
Content could be used, reused, and monetised by others, without recourse
Being sued is the least of your worries. It could mean losing your rights to content you've shared publicly. Once it's out there, it could be used, reused, and monetised by others, without recourse. Please read that again and let it sink in.
What You Can (and Can't) Protect
To make things simple, you can protect original work where you provided meaningful input, including AI-assisted content that you significantly shaped, edited, or curated.
You can also claim ownership of content generated through models you trained using your own datasets.
On the other hand, you can't protect work that's fully AI-generated without human contribution, content created with AI tools trained on copyrighted material without permission, or replicas of someone else's voice, style, or likeness used without their consent.
How to Protect Yourself
With all of this information in mind, it doesn't mean you should avoid using AI, far from it. By now, you know I'm an advocate of continuing to understand new technologies and how we can use them for good. But with AI, it does mean you need to be even smarter about how you use it.
Document your process. Keep a record of your prompts, revisions, and contributions. Show your creative input. It would be no different than back in the day when we needed to use pen and paper and use physical libraries and encyclopaedias to research and make notes.
Choose tools with clear terms. Work with platforms that disclose their training data sources and provide IP clarity. If it's vague, be cautious. It is easy to click "yes" to the terms and conditions box without reading what you are signing.
Don't imitate real people. As a professional, avoid generating content that mimics public figures, brand voices, or known artists unless you have the rights. There are already enough "15 minutes of fame" creators doing this. It is ok to get inspired by others and create fun viral content, but never lose your authenticity and respect copyrights.
Get legal guidance. If you're planning to scale or monetise AI-generated content, speak with someone who understands IP law in the AI space. Prevention is everything, and we are entering a world we still don't fully understand.
Stay educated. Copyright rules are evolving. What's legal today might shift tomorrow. Make it part of your creative discipline to stay updated. Websites such as the US Copyright Office or UK IPO AI and IP Consultation are good places to start.
Reflection time
If you're not thinking about this topic now, you may be forced to deal with it later. At best, you risk losing ownership. At worst, you face legal action, public backlash, and reputational damage.
Little story for you. A few years ago, before the current AI era, I used a royalty-free image (I had paid for the license) and was contacted by an agency claiming I was using the image without permission.
AI isn't going anywhere. It continues to evolve and challenge authorship, creativity, and ownership
They said I would have to pay a penalty unless I could prove I had legal rights. Since I had the license, it was not an issue, but it was a bit annoying to "waste" some of my time emailing back and forth.
And this was just for a licensed image from Getty Images, which I clearly had a license for. If content was tracked before, imagine now!
We are entering a new era (one where tools can create), but we're still accountable for what those tools do. And if we want to build something meaningful, something that lasts, we need to make sure it's ours to begin with.
For the gazillionth time, you'll hear me say, AI isn't going anywhere. It continues to evolve and challenge authorship, creativity, and ownership. But that doesn't mean we give up control. It means we get informed, build with integrity, and ask better questions.
If your next big idea comes from an AI tool, are you ready to claim it, protect it, and defend it if needed?
