AI vs Journalism: The New York Times considers legal action against OpenAI amid copyright tensions

Attorneys representing The New York Times are exploring the possibility of initiating legal action against OpenAI to safeguard the intellectual property rights associated with its journalistic content, according to confidential sources who shared the information with NPR.

Recent discussions between The Times and the developers of ChatGPT have been marked by tense negotiations aimed at reaching a licensing agreement. This agreement would involve OpenAI compensating The Times for incorporating its news stories into the artificial intelligence tools developed by the tech company. However, these negotiations have become increasingly contentious, prompting The Times to contemplate pursuing legal measures.

If The Times were to file a lawsuit against OpenAI, it would give rise to a significant legal clash centred on copyright protection within the realm of generative artificial intelligence.

A central concern for The Times revolves around the fact that ChatGPT is seemingly evolving into a direct competitor to the newspaper. This AI technology has previously generated text that responds to queries by drawing on original reporting and writing authored by The Times' staff. This apprehension is further fueled by the deployment of generative AI tools by technology corporations in search engines. For instance, Microsoft, a major investor in OpenAI, has incorporated ChatGPT into its Bing search engine.

The potential consequence is that users searching online might be presented with AI-generated responses that repackage content from The Times. This could significantly diminish the need for users to visit The Times' website, as noted by an individual involved in the ongoing negotiations.

Large language models such as ChatGPT gather extensive data from the internet to inform their responses to various queries. This data collection process often occurs without explicit permission. The legality of this practice remains an open question.

Should OpenAI be found guilty of copyright violations during this process, federal law permits the infringing materials to be eradicated at the conclusion of the legal proceedings.

In practical terms, if a federal judge determines that OpenAI unlawfully replicated articles from The Times to train its AI model, the court could order the destruction of ChatGPT's dataset. This would compel OpenAI to recreate its dataset solely using authorised content.

Federal copyright law also imposes significant financial penalties, potentially subjecting violators to fines of up to $150,000 per willful infringement.

The discussions between The Times and OpenAI have been ongoing in the wake of reports indicating that The Times would not join other media entities in endeavours to negotiate terms with technology companies concerning the utilisation of their content in AI models. A representative from The Times confirmed to NPR that this decision is not related to the potential litigation against OpenAI. OpenAI, on the other hand, declined to comment through a spokesperson.

Also Read Hollywood vs AI: Why famous actors including Oppenheimer, Barbie cast are on strike

Mounting Lawsuits

A potential lawsuit from The Times would join a series of legal actions recently directed at OpenAI.

Notably, comedian Sarah Silverman has joined a class-action lawsuit against OpenAI, asserting that the company utilised her 2010 memoir "The Bedwetter" without proper authorisation. She claims that the company accessed her work from an unauthorised online repository.

Other generative AI firms, including Stability AI, which operates the image generator Stable Diffusion, have also encountered copyright lawsuits.

Stability AI has been sued by Getty Images for allegedly training an AI model using over 12 million Getty Images photos without authorisation.

Legal experts anticipate that AI companies are likely to invoke the "fair use doctrine" as a defence. This doctrine allows for the utilisation of work without explicit permission under specific circumstances, such as for educational, critical, research, or news reporting purposes.

The ongoing AI copyright disputes are anticipated to hinge on two pertinent legal precedents.

Firstly, a federal appeals court ruling in 2015 determined that Google's digitisation of millions of books for its Google Books library qualified as legally permissible "fair use," rather than copyright infringement. The court's rationale was that Google's digital library did not constitute a significant market substitute for the original books, thereby not competing with the initial works.

Secondly, the Supreme Court's decision in the Andy Warhol Foundation case in May is expected to hold relevance for AI copyright disputes. In this case, the court determined that Andy Warhol's modification of a photograph taken by Lynn Goldsmith was not protected under the fair use doctrine. Importantly, the court observed that both Warhol and Goldsmith were profiting from the images by selling them to magazines.

Also Read

Battle of the billionaires: Elon Musk vs Mark Zuckerberg cage match could make over $1 billion

Google appeals to Supreme Court to quash antitrust directives on Android in India