OpenAI to Meta: AI companies that were accused of copyright infringement

The entire world was talking about generative AI when OpenAI launched ChatGPT in November 2022. The chatbot could write articles, summarise books, explain difficult terms, check grammar and spelling and a lot more. In a nutshell, activities that were previously considered exclusive to human beings were being carried out by the chatbot and it had initially left people surprised. As time passed, more AI chatbots like Gemini, Claude, MetaAI and CoPilot began emerging and people started using them for various purposes.

Now, we all know that AI is trained on huge sets of data. And this data, in some cases, was copyrighted. Laws regarding AI companies using copyrighted data are still blurry which is why such cases always make headlines. In the last two years, AI chatbots and companies behind them have faced several legal battles over copyright infringement, raising concerns about how these technologies are trained and used. From music and art to literature, content creators have increasingly accused AI firms of using their work without permission to train these systems. Let us take a look at five such cases.

1. Record Labels vs. Suno and Udio AI

In January 2024, the music industry’s powerhouses—Universal Music Group (UMG), Sony Music, and Warner Music Group took legal action against Suno and Udio AI, two emerging AI music companies. The record labels accused these companies of unlawfully using copyrighted music to train their AI systems. According to a report in Forbes, the crux of the lawsuit was in the claim that Suno and Udio AI used a vast amount of copyrighted sound recordings to enhance their AI's ability to generate music.

According to the labels, while the AI-generated music may not directly copy the original tracks, the extensive use of copyrighted material for training purposes constitutes a violation of intellectual property rights. The plaintiffs argued that the AI’s impressive ability to replicate musical styles and nuances was made possible by this unauthorised use of their recordings. The lawsuit sought not only financial damages but also injunctions to prevent further misuse of copyrighted music by these AI systems.

2. Scarlett Johansson vs. OpenAI

In May 2024, Hollywood actress Scarlett Johansson filed a lawsuit against OpenAI, alleging that her voice likeness was used without permission in an AI-generated voice model known as “Sky.” Johansson had previously declined OpenAI’s requests to use her voice for their chatbot, but the lawsuit claimed that her voice was still incorporated into the AI’s outputs.

The controversy began when Johansson’s legal team argued that despite OpenAI’s denial of using her voice, there were striking similarities between her voice and the AI-generated model. OpenAI maintained that the voice used was not Johansson’s but rather that of another actor. However, in response to the lawsuit, OpenAI decided to remove the “Sky” voice from their product out of respect for Johansson, despite continuing to defend their position.

3. Visual Artists vs. Google

In April 2024, a group of visual artists, including photographer Jingna Zhang and cartoonists Sarah Andersen and Hope Larson, brought a lawsuit against Google. A report by Reuters said that the artists accused the tech giant of using billions of copyrighted images to train its AI image generator, Imagen, without proper authorisation or compensation.

The lawsuit contended that Google’s AI was trained on a dataset comprising copyrighted artwork, which included the plaintiffs' works. The artists argued that Google had not obtained permission to use their images and that this unauthorised use was a breach of copyright laws.

On the other hand, Google denied the allegations, asserting that its models are trained on publicly available data.

4. Authors vs. Anthropic

Just last month, a class-action lawsuit was filed against Anthropic, the company behind the AI chatbot Claude. A report in The Guardian said that authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson accused Anthropic of using pirated copies of their books to train its AI model. The lawsuit alleged that Anthropic used these unauthorised copies to teach Claude how to generate text, effectively exploiting their intellectual property without permission.

The authors claimed that Anthropic’s actions not only infringed on their copyrights but also caused significant harm to their creative works. The lawsuit sought monetary damages and an injunction to prevent further misuse of the authors' works in training the AI.

5. Sarah Silverman and Authors vs. Meta

In December 2023, comedian Sarah Silverman, alongside other notable authors like Pulitzer Prize winner Michael Chabon, sued Meta Platforms (formerly Facebook). The lawsuit accused Meta of using pirated copies of their books to train its Llama AI language model. Despite legal warnings from its own team, Meta allegedly proceeded with using these unauthorized copies, which led to the lawsuit.

The plaintiffs argued that Meta’s actions were a blatant disregard for intellectual property laws and ethical standards. They sought damages and an order to prevent Meta from using pirated content in its AI training practices.

SEE ALSO:
Apple's iOS 18 is coming soon: List of compatible devices, key features
The internet has more AI-generated content today than human-created and this could mean trouble