Google AI rejects requests that mention Disney characters

Brief on the Disney–Google conflict

In December 2023, Disney accused Google of using its AI models (Gemini and Nano Banana) to create images of characters from its own franchises without permission. According to Disney, these models turn user requests into “virtual vending machines,” generating high‑quality pictures of Yoda, Darth Vader, Iron Man, and Winnie the Pooh.

Since then, Google disabled the ability to generate images of those characters. When a user attempts to request any of them via text, the model responds:

> “I’m unable to generate the requested image right now due to concerns from third‑party content providers. Please edit your query and try again.”

However, experts note that if users upload photos of the characters along with textual prompts, Google’s AI can still produce related images.

How the conflict began
1. December 2023 – Disney sent Google an official letter (32 pages) demanding a halt to copyright infringement and model training on its content.
2. The letter provided specific examples of simple requests to Veo, Nano Banana, and Gemini that produced high‑quality Disney character images.
3. For several months before the formal step, Disney tried to express concern but without success.

Google’s response
A company representative said:

> “We use publicly available data from the internet to train our AI and have developed additional copyright control tools (Google‑extended, Content ID for YouTube) that allow sites and rights holders to manage their content.”

Context
- The Disney letter was sent almost simultaneously with Google’s $1 billion licensing agreement with OpenAI.
- In its statement, Google emphasized long‑term, mutually beneficial relationships with Disney.

Thus, the conflict arose because Google’s AI technologies were generating copyrighted characters without the owner’s permission. After Disney’s formal complaint, the model was filtered, but questions about content control remain open.