US urges G20 to allow AI training on creators' work

The United States has urged G20 nations to establish guidelines governing how artificial intelligence companies utilize copyrighted material for training their systems, amid a surge in lawsuits from American creators. Tech giants such as Anthropic, OpenAI, Google, and Meta are currently embroiled in legal battles with authors, publishers, media organizations, and other creators who claim their works were improperly used to develop AI products.

These companies employ AI models to generate images and text by ingesting vast quantities of data, including books, music, and art. They argue that employing this data to build their AI tools is permissible under the principle of fair use.

U.S. Commerce Secretary Howard Lutnick conveyed this stance to G20 officials during a technology-focused meeting in North Carolina. He proposed that countries should permit AI firms to train their models on creators' work while simultaneously safeguarding the rights of artists. Lutnick emphasized the necessity of frameworks that protect innovators without impeding the progress of future innovators. He did not specify how this could be achieved.

The U.S. Justice Department concurrently filed a brief supporting OpenAI in its lawsuit against the New York Times and several other newspapers for utilizing their work to train the language models powering ChatGPT. The department argued that AI training typically constitutes fair use of copyrighted material. Nvidia CEO Jensen Huang joined Lutnick prior to Lutnick's speech, urging G20 officials to steer clear of regulations based on hypothetical hazards and instead develop rules addressing tangible issues related to the technology.

This sentiment mirrored that of the Trump administration and other American tech executives who addressed the gathering on Tuesday.

The G20 nations aim to avoid crafting new regulations exclusively for AI, according to a joint statement released by the U.S. Any forthcoming rules should address only new considerations arising from AI that are not currently covered by existing laws, the statement stated. Government regulations could impede the industry's profitability by delaying the release of new models or compelling companies to alter their product performance for security reasons.

China, which has instituted a comprehensive array of AI regulations, was among the signatories to this statement. Additionally, prominent AI industry figures, including Anthropic co-founder Tom Brown, Palantir's Alex Karp, and OpenAI's Sam Altman, addressed the gathering on Wednesday morning.