The claim, first reported by Variety, pushes one of Hollywood’s most powerful studios directly into the center of the AI training debate—where tech companies, creators, and rights holders are still fighting over what “fair use” really means in the age of generative AI.
What Disney says Google’s AI crossed the line on
According to Disney’s legal team, Google’s AI systems have been trained on — and are now generating — unauthorized images and videos featuring some of the company’s most valuable intellectual property.
The list reads like a greatest-hits catalog: Frozen, The Lion King, Moana, The Little Mermaid, Marvel titles like Deadpool and Guardians of the Galaxy, and franchises including Star Wars.
Disney also raised a more subtle but serious concern: some AI-generated outputs reportedly include Google’s own logo. That combination, Disney argues, could mislead audiences into believing the content is officially licensed or approved by Disney itself.
Months of private talks, then a legal warning
This wasn’t Disney’s opening move. The company says it had been raising concerns with Google for months, attempting to resolve the issue behind closed doors.
When those discussions failed to produce a satisfactory outcome, Disney escalated with a formal cease-and-desist—essentially putting Google on notice to stop the alleged misuse of its copyrighted material.
It’s a familiar pattern in the AI era: quiet negotiations first, public pressure next.
Google responds — and starts taking content down
Following the notice, Google began working with Disney to remove dozens of AI-generated videos featuring Disney characters, according to Deadline.
At the same time, Google pushed back on the broader claim. A company spokesperson reiterated that Google trains its AI models using “public data from the open web,” a position shared by many AI developers.
Google also pointed to its “longstanding and mutually beneficial relationship” with Disney, highlighting tools like Content ID, which help rights holders identify and manage copyrighted material across platforms.
Still, the core tension remains unresolved. AI models are trained on enormous datasets scraped from the web, and copyrighted material often enters those datasets quietly—until recognizable characters like Mickey Mouse or Elsa start appearing in generated content.
Disney isn’t anti-AI — it just wants control
Perhaps the most revealing detail came the same day Disney took legal action against Google.
The company finalized a $1 billion licensing and investment deal with OpenAI, granting OpenAI’s Sora video generator permission to create content using more than 200 Disney, Marvel, Pixar, and Star Wars characters over the next three years.
Some of that AI-generated, user-created content is expected to appear on Disney+, making it clear that Disney isn’t rejecting AI outright. Instead, it’s drawing a firm distinction between licensed AI use and unapproved training or outputs.
Why this clash matters beyond Disney and Google
This dispute highlights a widening gap in how tech companies and content owners view AI training. For AI developers, publicly available data is fair game unless explicitly restricted. For studios like Disney, availability doesn’t equal permission—especially when the output can compete with or dilute the original brand.
The fact that Disney is simultaneously suing one AI giant while partnering deeply with another underscores where the industry may be headed: not blanket bans, but tightly controlled licensing deals.
As courts, regulators, and companies continue to wrestle with these questions, one thing is becoming clear—AI may be moving fast, but legacy media companies are no longer willing to sit on the sidelines.
The question now: will future AI innovation depend less on scraping the open web, and more on who can secure the biggest licensing deals?
