AI has been facing a significant copyright issue for a while. However, xAI’s Grok 2 has reached a new level. The AI startup recently added a new image generator called Aurora to the chatbot. As a result, X users have inundated the platform with AI-generated pictures of celebrities. While some are experiencing issues with their requests, others have created images of famous people in compromising situations.
xAI CEO Elon Musk has referred to Aurora as an uncensored AI model. However, the tool is raising questions about whether it’s functionality is legal or not.
Breaking the rules. Unlike other AI tools, Grok allows for the straightforward generation of celebrity images. OpenAI has faced several lawsuits for using copyrighted content. However, Musk seems to be following the adage of “asking for forgiveness, not permission.”
Scarlett Johansson’s warning. The issue of using copyrighted images and content for AI training has been debated for years, with actress Scarlett Johansson being one of the most vocal opponents. OpenAI had to remove a voice that closely resembled hers. The company claimed it was “never intended to resemble hers” but made the change to avoid legal repercussions.
The similarities between Grok’s generated images and the likeness of celebrities are also quite apparent.
How did xAI train Aurora? X is filled with celebrity photos. xAI engineers have used this large amount of photos shared on the platform to train Grok’s AI model. In the summer, X automatically enabled all users to allow their data to be used for AI training. While it was possible to opt out, millions of users likely didn’t do so. Others might have been unaware of the feature.
While some image generators rely on large public image banks, Grok taps into a unique resource: a proprietary database accessible only to X itself.
Is this legal? Musk seems to know what he’s doing. The issue is that there’s no established legal framework in the U.S. regarding the use of royalty-free images for training AI.
Musk’s defense rests on the concept of “fair use,” which allows certain uses of copyrighted materials without needing permission from the copyright holder. According to X’s platform use guidelines, “Certain uses of copyrighted material may not require the copyright owner’s permission. In the United States, this concept is known as fair use. Some other countries have a similar concept known as fair dealing.”
While many entities in the industry are aware that Musk’s use of AI may not be legal, a definitive ruling from the Supreme Court could take years. In the meantime, Grok is achieving results that its competitors can’t match.
The responsibility for any misuse rests with the user. xAI’s Terms of Service say, “We take no responsibility and assume no liability for any input. You will be solely responsible for your input and the consequences of submitting, posting, displaying, providing, sharing, or otherwise making it available on or through the service.”
The company also clarifies what users can’t do with Grok: “You must not use any output relating to a person for any purpose that could have a legal or material impact on that person, such as making credit, educational, employment, housing, insurance, legal, law enforcement, citizenship, medical, or other important decisions about them.”
Experts in digital law have noted that criminal images can be reported regardless of whether they were created with AI or drawn by hand. They also point out that responsibility for the content lies with the person who publishes it, not the AI tool or the social media platform.
Will any celebrities file lawsuits against xAI? The tool has been available for a few days. However, there’s no record of any legal actions filed against xAI or X. A lawsuit would be necessary for the U.S. justice system to determine whether the use of copyrighted images falls under fair use.
Many celebrities are particularly sensitive about others profiting from their image, which could also lead to a lawsuit against the platform. This legal battle would likely be complex and could only be pursued by high-level law firms.
Europe’s million-dollar sanctions. Grok is yet another reason that the European Commission is considering imposing substantial sanctions on X. The Commission already warned Musk that his platform had violated the Digital Services Act after it found “motivated malicious actors abusing the ‘verified account’ to device users.” As such, X could face fines of up to 6% of its annual revenue. Musk said in July that he intends to fight this in court.
Under the Digital Services Act, X is obligated to remove illegal content and provide mechanisms for users to report it. In this context, the concern regarding Grok revolves around images of celebrities in situations that could damage their reputation.
xAI’s other legal challenges. xAI faces another legal battle related to the training of its image generation model. While this use might be covered by fair use in the U.S., using copyrighted images to train AI is prohibited under the new European Union AI Act.
More than 100 companies, including Google, Amazon, OpenAI, and Microsoft, have agreed to comply with the European regulation. However, xAI isn’t on that list. Musk seems to be doing its own thing until a court rules otherwise, which could take considerable time.
Image | Rock History/X| Emad/X
View 0 comments