In 1956 the Federal Communications Commission ordered AT&T to license its 7,820 unexpired patents, including those held by Bell Labs, to any American firm on reasonable terms. The order followed a seven‑year negotiation between AT&T and the federal government and was intended to curb the company’s vertical integration. AT&T had operated as a regulated monopoly with a stable revenue base that funded extensive research at Bell Labs, resulting in numerous scientific breakthroughs and a portfolio of patents that later generated an estimated $6 billion in downstream value. The settlement allowed firms such as Intel to commercialize technologies that originated from Bell Labs, contributing to the growth of the U.S. semiconductor industry. Today, large language models are trained on extensive collections of publicly available internet text, code, and other content. Companies such as OpenAI and Anthropic assert that this use is transformative and falls under fair use, while rights holders argue that the training process reproduces protected works without permission. In June 2025 Judge Alsup ruled in Bartz v. Anthropic that training on lawfully acquired books was transformative, but training on pirated books constituted infringement, leading Anthropic to settle for $1.5 billion. A separate ruling in Kadrey v. Meta found training on publicly available data to be transformative, though the court found insufficient evidence of market harm. The U.S. Copyright Office has stated that public availability does not automatically grant fair‑use rights, and no definitive legal standard for AI‑related market dilution exists. Some analysts propose a “corpus royalty” in which AI developers would contribute a fixed share of revenue to a public fund distributed equally among U.S. residents. The proposal is intended to compensate the collective contribution of internet users whose content fuels model training and to align incentives between public contributors and private AI enterprises. The author notes that without mechanisms to ensure a steady flow of diverse user‑generated content, the internet ecosystem that underpins AI progress may degrade, and that a royalty system could help restore balance.