Stephen Thaler, a computer scientist, filed two US patent applications naming an artificial intelligence system he built - DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) - as the sole inventor. The claimed inventions were a fractal-surfaced beverage container and a flickering emergency light. The US Patent and Trademark Office rejected the applications because no human was listed as an inventor, and Thaler challenged that refusal in court.
On August 5, 2022, the US Court of Appeals for the Federal Circuit affirmed the USPTO’s denial in Thaler v. Vidal. The court held that the Patent Act requires an inventor to be a “natural person,” reasoning that although the statute does not define “individual,” the ordinary meaning of that word - reinforced by Supreme Court precedent and the Act’s use of pronouns like “himself” and “herself” - is a human being, not a machine. The court deliberately set aside the philosophical question of whether AI can truly invent, deciding only the narrow statutory question of who can be named. The Supreme Court declined to hear Thaler’s appeal in 2023, leaving the ruling in place.
Thaler ran parallel DABUS cases worldwide, with similar results. The United Kingdom, the European Patent Office, and others also held that an inventor must be human, though South Africa granted a patent and an Australian court initially sided with Thaler before being reversed on appeal.
Why business readers should care: as companies increasingly use AI in research and development, this case settled a foundational question - a machine cannot hold inventorship. Patents on AI-assisted inventions must name the humans who contributed, which has direct consequences for how firms document and claim their AI-driven innovations.