On March 2, 2026, the U.S. Supreme Court declined to hear a case challenging the principle that AI-generated works, without human authorship, cannot be copyrighted, according to Fadel. This decision upheld lower court rulings in Thaler v. Perlmutter, solidifying the legal position: content created solely by algorithms lacks federal protection. Creators relying exclusively on generative AI face a significant challenge in safeguarding their output.
Content creators rapidly adopt AI tools, but the legal framework for protecting AI-generated content remains firmly rooted in human authorship. This creates a critical ownership gap, leaving many works vulnerable to uncompensated reuse and infringement. The U.S. Copyright Office confirms that AI-generated content alone cannot receive copyright protection, as detailed by Spencer Fane.
Therefore, companies and individual creators must adapt workflows to ensure demonstrable human contribution in AI-assisted creations. Otherwise, they risk losing exclusive rights and control over their digital assets. While AI platform Terms of Service may grant contractual ownership, this does not guarantee federal copyright protection against third-party infringement; it cannot supersede federal law.
This legal landscape creates a paradox: contractual rights from AI platforms offer no real-world protection against third-party infringement. A platform's Terms of Service (TOS) might state users own generated content, but this is a contract between user and platform, not a federal copyright. This distinction is crucial, as federal copyright provides robust legal standing against unauthorized use. Some AI platforms also reserve rights to reuse or license generated content, especially for free-tier plans, as noted by Fadel. Creators could unknowingly sign away control over their 'work' without gaining federal copyright. Without human authorship, the content is effectively public domain, regardless of platform agreements.
The Dual Challenge: AI's Use of Copyrighted Works and the Expanding DRM Market
In Bartz v. Anthropic PBC, the court ruled that training large language models (LLMs) with copyrighted books constituted fair use, finding the use 'exceedingly transformative'. However, this decision did not grant a blanket exemption for all AI data acquisition. The court denied Anthropic's motion for summary judgment on the plaintiffs' piracy claim, finding that knowingly downloading and copying pirated works was not fair use.
