Tracy Pratt, Brandfully Yours columnist
Dear Favorite Business Leader: A year ago, the AI legal storm was still brewing. Lawsuits were filed, arguments staked out, but almost nothing was decided. This summer, several of those cases finally reached real outcomes, and the results are beginning to reshape how businesses need to think about the AI tools they use.
Here’s what actually landed, what’s still moving and why it matters for your business (with the caveat that a couple of these could shift before publication).
Price of pirated training data
The largest of these cases is now closed. On July 20, a federal judge granted final approval to Anthropic’s $1.5 billion settlement with a class of authors and publishers in the largest known copyright settlement in U.S. history.
It resolves claims that Anthropic had built a research library from more than 7 million pirated books. It’s worth noting that an earlier ruling in the same case still stands, stating that training AI on legally acquired books can qualify as “fair use” (that’s the rule that lets you legally use copyrighted work without permission for things like criticism, teaching and news).
It wasn’t the training itself that cost Anthropic, it was the piracy.
The implication for business: If you’re using a vendor’s AI tools, it’s worth asking how their models were trained, on what data and whether they’re licensed to use that data.
Human authorship required
In March, the U.S. Supreme Court declined to hear Thaler v. Perlmutter, leaving in place a rule that work generated entirely by AI, with no meaningful human input, cannot be copyrighted. You might remember this one. It stems from a 2018 attempt to copyright an artwork created by AI.
The implication for business: If your team is using AI to produce code, marketing copy or designs with minimal human editing, that work may not be legally protected, which means a competitor could copy it freely. So, make sure a human is the one shaping and revising anything you want to copyright.
One to watch: NYT v. OpenAI
Of all these cases, this is probably the one worth watching most closely.
In July, the Times and a group of other publishers accused OpenAI of concealing its ability to search training data and chat logs for their copyrighted journalism and asked the court to sanction the company over it. That motion was still pending as I wrote this, so a ruling could land before this publishes. The underlying question: Is training a commercial AI model on someone else’s journalism “fair use,” or is it theft?
The implication for business: If the court rules that training commercial AI models on copyrighted journalism is not fair use, then the business model of generative AI providers might pivot toward expensive, permission-based commercial licensing, and that might affect the free chat queries we’ve all enjoyed.
We’ve graduated from AI chatbots to AI agentsTrademark, not copyright, may be AI’s sharper exposure
Getty Images’ fight with Stability AI is really two separate cases with two different outcomes so far.
In the U.K., a court largely rejected Getty’s claims in November, though Getty has since won permission to appeal.
In the U.S. case, refiled in Northern California, a judge allowed Getty’s trademark and unfair-competition claims to move forward — not because Stable Diffusion’s AI copied Getty’s photos outright, but because it kept generating images with Getty’s watermark on them, distorted but recognizable.
The implication for business: If you’re running an internal AI tool and it starts producing content with a competitor’s logo baked in, even a garbled version, you may have a trademark problem on your hands before copyright even enters the conversation.
Germany’s warning shot on AI liability
The case I find most interesting is unfolding in Munich.
In late May, a German court issued an injunction against Google over its AI Overviews feature, after the AI “hallucinated” a false story linking a small publisher to a scam and subscription-trap schemes. The court’s reasoning was direct: an AI-generated summary is Google’s own statement, not a neutral list of links, and Google doesn’t get to point to the AI as a separate actor (The Decoder).
Google is appealing, so nothing is final yet. But that same argument — that an AI provider is the author of its own output — might eventually force American courts to confront Section 230 of the Communications Decency Act. That’s the law that has long shielded digital platforms from liability for what their users post.
The implication for business: If U.S. judges were to reach a similar conclusion, and I’m not a lawyer, but I’m guessing that might mean a business could potentially be held liable if it runs a chatbot or auto-generated summary that makes a false statement about a person or a competitor.
Also on the radar
Two more worth a mention.
In May, a jury took less than two hours to reject Elon Musk’s $150 billion suit against OpenAI on a technicality — he’d waited too long to sue. Musk says he’s appealing, so this one isn’t yet fully closed.
And in July, Apple sued OpenAI, accusing former Apple employees of taking trade secrets with them to build OpenAI’s new hardware line.
Neither case touches most businesses directly, but the Apple suit is a good reminder to revisit your own onboarding and offboarding practices. As AI hiring wars heat up, this kind of exposure is likely going to get more common.
Key takeaways for your businessVet your AI vendors. Ask where their training data came from and whether it was properly licensed. The era of AI companies training first and asking questions later appears to be closing fast, and the businesses relying on those tools could inherit some of that risk.Keep humans in the loop. If you want to legally own something an AI platform helped create, make sure a human is meaningfully editing and shaping the final product.Watch your outputs. If your AI tools ever generate content featuring a competitor’s logo, watermark or branding, treat that as a flag worth investigating.Budget for licensing, not just tools. If courts start ruling against fair use for training on journalism and other copyrighted content, expect AI providers to pass those licensing costs down the line. It wouldn’t hurt to build some flexibility into your future AI budgets.Tighten your IP paperwork. As AI talent wars heat up, so does the risk of incoming hires bringing in a former employer’s trade secrets. A quick review of your onboarding and offboarding protocols could be some cheap insurance.Don’t bank on Section 230 to cover your own AI. If that Munich reasoning holds up here, a false or damaging statement your chatbot generates could be treated as your business’ own words, not a neutral post you’re shielded from, the way you would be for something a customer or user wrote.
This legal storm hasn’t finished moving through yet. If anything, we’re just getting into the thick of it. But you don’t want to wait for the sky to turn derecho yellow to figure out where the flashlights are (yeesh, that was cheesy!). Anyway, a little preparation now can go a long way.
Brandfully yours,
Tracy