The landscape of artificial intelligence development is facing increasing scrutiny over the use of copyrighted material in training models. OpenAI, the developer behind the groundbreaking chatbot ChatGPT, finds itself at the center of controversy amid accusations of “unlawful use” of copyrighted work by the New York Times and other legal complaints.
OpenAI recently submitted a statement to the House of Lords communications and digital select committee, emphasizing the crucial role of copyrighted material in training large language models. The company contends that the vast array of data, encompassing blog posts, photographs, forum discussions, software code, and government documents, falls under copyright protection. OpenAI asserts, “Because copyright today covers virtually every sort of human expression, it would be impossible to train today’s leading AI models without using copyrighted materials.”
The developer addresses the limitations of relying solely on out-of-copyright books and drawings, stating that such constraints would result in inadequate AI systems for meeting contemporary citizens’ demands. OpenAI contends that the legal doctrine of “fair use” permits the use of copyrighted content under specific circumstances without seeking explicit permission. The company firmly believes that copyright law does not forbid the training of AI models.
In response to the New York Times lawsuit, OpenAI expressed its support for journalism, collaborated with news organizations, and dismissed the lawsuit as lacking merit. The company reiterates its respect for the rights of content creators and owners. Despite the legal challenges, OpenAI emphasizes its commitment to AI safety and backs independent analysis of its security measures. The submission advocates for “red-teaming” of AI systems, where third-party researchers emulate the behavior of rogue actors to test the safety of the products.
OpenAI is not alone in facing legal action, as other AI companies, including Amazon-backed Anthropic and Stability AI, are entangled in copyright infringement lawsuits. The legal landscape surrounding the use of copyrighted material in AI development is becoming increasingly complex, prompting discussions on the balance between innovation and intellectual property protection.
In essence, OpenAI finds itself navigating a challenging terrain where the necessity of copyrighted content for training AI models clashes with legal disputes and calls for enhanced safety measures in the rapidly evolving field of artificial intelligence.