This Legal Battle Could Reshape AI

Not long ago, I wrote about how artificial intelligence is changing the rules of the internet.

Instead of simply pointing people toward information with Google Search, the company now wants to answer our questions “in-house” using AI.

Today, I want to take that idea one step further. Because it turns out that changing the rules of the internet may also mean changing the rules that govern it.

Last week, a court in Germany ruled that Google could be held liable for false statements generated by its AI Overviews.

And I believe that decision could have consequences that extend far beyond Google.

Search Engines vs. AI

For decades, courts have generally treated search engines like librarians.

The job of search wasn’t to create information, it was to organize it.

That meant, if someone wrote something false on a website, Google wasn’t considered the author. After all, Google Search simply helped users find that page.

But Google’s AI Overviews work differently.

Instead of presenting a list of links, it reads information from multiple sources, combines it and generates an entirely new response.

That’s exactly what got the company into trouble in Europe.

According to the lawsuit, Google’s AI accused two publishers of being connected to scams.

But nobody had actually written that. Instead, the AI read information from several different websites, blended it together and generated a false statement on its own.

Maybe you’ve experienced hallucinations like this while working with AI. I know I have.

And according to the court, that’s exactly what makes this situation different from traditional search.

Google’s AI didn’t simply help people find information. It created information that wasn’t there. And the judges ruled that Google could be held responsible for it.

To understand why this could become such a big deal, you have to understand one of the most important laws in internet history.

It’s called Section 230.

Passed in 1996, Section 230 helped create the modern internet. In simple terms, it says online platforms usually aren’t legally responsible for what other people post.

So if someone writes something defamatory on a message board, the person who wrote it can be sued. But the platform usually can’t.

That protection gave companies like Google, Facebook, YouTube and X room to grow into massive businesses.

But Section 230 was written long before generative AI existed.

Back then, websites mostly displayed content created by other people. Now AI systems increasingly generate content of their own.

That’s exactly what the German court focused on.

The judges rejected Google’s argument that a simple disclaimer is enough to warn users that AI can make mistakes. Instead, they concluded that Google’s AI had produced its own false statements.

If other courts reach the same conclusion, it could force them to rethink how existing internet laws apply to generative AI.

America Is Still Deciding

Here in the United States, judges are only beginning to wrestle with these questions.

So far, the legal battle over who should be responsible for AI-generated content is mostly playing out in lower courts and state lawsuits.

In Georgia, OpenAI defeated a defamation lawsuit after ChatGPT falsely accused radio host Mark Walters of financial misconduct. The court found Walters had not shown enough evidence to hold OpenAI legally responsible.

That was a win for AI companies.

But another case is moving in the opposite direction.

In Florida, a federal judge allowed a wrongful death lawsuit against Character.AI to move forward after a teenager died by suicide following months of conversations with one of its chatbots.

Character.AI argued that chatbot output should receive First Amendment protection. But the judge wasn’t ready to go that far.

Meanwhile, Pennsylvania has sued Character.AI for allegedly allowing chatbots to present themselves as licensed doctors. Investigators reportedly found AI characters claiming to be medical professionals and giving health-related guidance.

So this isn’t just about random hallucinations.

Courts and regulators are finally beginning to ask whether AI companies are responsible for the systems they release into the world.

That includes what those systems say, what they recommend and how they interact with children.

It also raises questions about whether users can trust them in high-stakes situations like health care or financial advice.

And that could change the next phase of AI development.

For the past few years, the industry has raced to build models that are faster, cheaper and more powerful. But companies may soon also need to consider how to make their AI systems explain where their answers came from.

That would require stronger safeguards, more reliable source tracking and clearer warnings when an answer is uncertain.

As I noted in my earlier piece, Google says that AI Overviews now reach roughly 2.5 billion users. Gemini has grown to around 900 million monthly users. And AI Mode has already surpassed 1 billion users.

Chart: Google

And now that these products reach billions of people, even small failure rates have the potential to become an enormous legal risk.

Here’s My Take

The questions surrounding AI aren’t just about what it can do, but about who is responsible when it gets things wrong.

That means liability could shape how AI products are built. It could also influence how quickly companies roll out new AI features

And as I’ve been saying for a while, it could create demand for an entirely new layer of AI infrastructure built around trust, verification and safety.

Because once AI starts creating information instead of simply finding it, the rules change.

And courts are only beginning to catch up.

Regards,

Ian KingChief Strategist, Banyan Hill Publishing

Editor’s Note: We’d love to hear from you!

If you want to share your thoughts or suggestions about the Daily Disruptor, or if there are any specific topics you’d like us to cover, just send an email to [email protected].

Don’t worry, we won’t reveal your full name in the event we publish a response. So feel free to comment away!