AI models such as ChatGPT, Claude, and Gemini have gained substantial knowledge about the world by being trained on extensive datasets sourced from the internet. However, not all the information utilised in this training was obtained with permission or under fair terms. Allegations have arisen against organisations like OpenAI and Anthropic for allegedly using copyrighted material from authors, publishers, and news outlets without consent or remuneration.

Brent Hecht, Microsoft’s director of applied science, described this situation as potentially the largest theft of labour in history. His comments emerged from newly unsealed court documents related to a copyright lawsuit filed by The New York Times and various other news organisations against Microsoft and OpenAI. Hecht argued that individuals around the globe would likely view the practices of AI scraping their creative contributions as a monumental theft.

Hecht’s deposition highlighted that the majority of content creators did not intend for their works to be utilised in such a manner and expressed concern that they were not compensated for their intellectual property. This brings to light the ethical dilemma surrounding the usage of online content for AI training.

Fair Use Defence in Copyright Case

Both Microsoft and OpenAI contend that their utilisation of copyrighted material falls under the fair use doctrine. They argue that their AI systems transform the original content substantially rather than merely reproducing it. This position has been contested by the news organisations involved in the lawsuit, which claim that the AI-generated outputs can replace the need for direct engagement with the original news sources.

Internal communications reveal insights into how AI chatbots could reshape news consumption. For example, individuals could obtain information directly through a chatbot instead of visiting a news website, which may impact traffic to these sites. Microsoft CEO Satya Nadella acknowledged that engaging with chatbots has replaced the traditional process of sourcing information from news websites for users.

The filing indicates that this development poses challenges for publishers, as AI-generated models utilise content created by news outlets while simultaneously decreasing the necessity for users to directly visit these digital platforms. Data cited in the filings indicated that click-through rates on platforms like The New York Times and Daily News saw a significant drop, ranging from 83 to 93 per cent lower when engaging with Microsoft’s “answer engine” compared to traditional Bing Search.

Implications for Future News Publishing

The unfolding lawsuit, initiated by The New York Times in 2023 soon after the launch of ChatGPT, sparked further inclusion of additional news organisations, such as the New York Daily News, The Intercept, and the Center for Investigative Reporting, seeking to uphold their copyright claims. Microsoft and OpenAI have continuously maintained their fair use argument, asserting the transformative nature of their AI products.

Comments from key figures within the companies suggest a focused effort on utilising AI models for news content effectively. OpenAI co-founder Greg Brockman noted that AI models exhibit strong performance in predicting the text of news articles, indicating that they are well-suited for tasks related to news production.

As this legal battle progresses, the court will ultimately need to assess the validity of the news organisations’ copyright claims against the backdrop of emerging AI technologies and their implications for content creators. The outcome could have far-reaching effects on the future relationship between AI development and traditional publishing entities.

Long or Short, get news the way you like. No ads. No redirections. Download Newspin and Stay Alert, The CSR Journal Mobile app, for fast, crisp, clean updates!

App Store –  https://apps.apple.com/in/app/newspin/id6746449540 

Google Play Store – https://play.google.com/store/apps/details?id=com.inventifweb.newspin&pcampaignid=web_share