Public availability shields the company from liability, Apple argues
In a courtroom that may quietly redraw the boundaries of creative ownership in the digital age, Apple has moved to dismiss a lawsuit brought by three YouTube creators who allege their videos were harvested without consent to train artificial intelligence systems. The company's defense rests on a deceptively simple premise: that which is publicly visible is publicly usable. This case joins a growing constellation of legal disputes asking whether the internet's openness is a license or merely an invitation — and who, ultimately, holds the rights to the raw material of machine intelligence.
- Three YouTube creators discovered their publicly posted videos had been quietly absorbed into Apple's AI training pipeline without their knowledge or compensation.
- Apple responded not with settlement talks but with a motion to dismiss, arguing that public availability of content dissolves the barrier between access and permission.
- The DMCA — a law designed to protect against circumvention of digital locks — is being invoked by Apple to argue there were no locks to circumvent, and therefore no violation.
- Creators and rights advocates warn that this logic, if accepted, would effectively strip online publishers of any meaningful control over how their work is commercially exploited.
- The court must now decide whether to let the case proceed to discovery, where Apple's actual scraping practices would face scrutiny, or to end it here on legal grounds alone.
- A ruling either way could become a landmark precedent governing how the entire AI industry sources the human creativity it depends upon.
Apple has filed a motion to dismiss a copyright lawsuit brought by three YouTube creators who allege the company scraped their videos without permission to train its artificial intelligence models. The creators discovered their content had been harvested for machine learning purposes and argued this constituted unauthorized commercial exploitation of their work — a claim Apple is now asking a federal court to reject outright.
Apple's defense turns on a single foundational argument: the videos were posted publicly on YouTube, making them freely accessible to anyone, including Apple, without circumventing any technical barrier. The company invoked the Digital Millennium Copyright Act to contend that its protections simply do not apply when no encryption or access control was ever in place. In Apple's framing, there was nothing to unlock — and therefore nothing unlawful in the taking.
This argument mirrors a posture that has become common across the technology industry, where public availability is routinely treated as implicit permission. Creators and rights holders have pushed back forcefully against this interpretation, insisting that posting work online does not constitute a blanket license for any subsequent commercial use. The gap between those two positions is precisely what courts are now being asked to resolve.
If Apple's motion succeeds, the creators would need to appeal or pursue the case under different legal theories. If it fails, the lawsuit advances to discovery — a phase that could expose the full scope of Apple's data collection practices. Either outcome carries weight well beyond this single dispute, potentially setting the terms under which AI companies may legally harvest the creative output of the open internet.
Apple filed a motion to dismiss a lawsuit brought by three YouTube creators who accused the company of scraping their videos without permission to train artificial intelligence models. In its defense, Apple argued that the videos in question were publicly available content, and that this fact shields the company from liability under the Digital Millennium Copyright Act—a federal law that generally protects circumvention of copyright protections, but with exceptions for lawful access to publicly available material.
The three YouTube channels initiated legal action after discovering that Apple had harvested their video content to feed into machine learning systems. The creators contended that this use violated their copyright protections and amounted to unauthorized commercial exploitation of their work. The case touches on a fundamental tension in modern AI development: the hunger for vast training datasets versus the rights of creators whose work becomes fuel for those systems.
Apple's legal strategy hinges on a straightforward claim: because the videos were posted publicly on YouTube, anyone—including Apple—could access them without breaking any law or circumventing any technical barrier. The company invoked DMCA provisions to argue that even if copyright infringement occurred, the statute's framework does not apply when the material in question was never locked behind encryption or access controls. In other words, Apple contends there was nothing to circumvent.
This defense reflects a broader pattern in how technology companies have approached AI training data. The argument that public availability equals permissible use has become standard in the industry, even as creators and rights holders push back against the notion that posting something online amounts to granting a license for any subsequent use. The distinction between access and permission remains contested legal territory.
The lawsuit represents one of several challenges facing Apple and other major tech firms over their data sourcing practices for AI systems. As these companies race to build and refine large language models and other AI tools, questions about the legality and ethics of their training methods have intensified. Courts will ultimately need to decide whether publicly available content can be freely harvested for commercial AI training, or whether creators retain rights that prevent such use without consent.
Apple's motion to dismiss asks the court to throw out the case before it reaches trial, arguing that the plaintiffs' claims fail as a matter of law. If the motion succeeds, the YouTube creators would need to appeal or refile under different legal theories. If it fails, the case would proceed to discovery and potentially trial, where both sides would present evidence about Apple's scraping practices and the scope of copyright protections for online video content. The outcome could shape how tech companies legally source training data for AI systems and what constitutes permissible use of publicly available creative work.
Citações Notáveis
Apple argued that videos used for AI training were publicly available content protected under DMCA provisions— Apple's legal filing
A Conversa do Hearth Outra perspectiva sobre a história
Why does Apple think being public changes anything about copyright?
Because they're arguing there's no legal barrier to access. If you posted something on YouTube, anyone can watch it. Apple says that means they didn't break any law to get it—they just watched, like anyone else could.
But watching and copying for AI training are different things, aren't they?
Exactly. That's the real fight. Apple says the DMCA doesn't apply because there was nothing to circumvent. The creators say copyright still protects how their work gets used, regardless of whether it was behind a paywall.
So what does the DMCA actually protect here?
It's supposed to prevent people from breaking through technical locks to access protected content. Apple's argument is: there were no locks, so the DMCA doesn't matter. But the creators' lawyers probably say copyright is separate from the DMCA—you can't use the DMCA as a shield for copyright infringement.
If Apple wins this motion, what happens?
The case gets dismissed before trial. The creators would have to appeal or find a different legal angle. If they lose, it signals that public availability is a strong defense for AI training data scraping.
And if they win the motion?
Then Apple has to actually defend itself in court, prove its scraping was legal, and face a jury or judge deciding whether public posting really does mean permission to train AI.