Who Owns AI Generated Inventions: Machine Creativity and Patent Ownership

69 questions on who owns AI generated inventions, answered and cited by the UpLaw editorial team.

• Are correct inventorship and candor competing obligations?
• Does a composition claim face the same eligibility problem as a method claim?
• Does Pannu still apply to anything after the 2025 guidance?
• Does PERA address AI inventorship?
• Does Section 112 bite differently for a composition than for a model?
• How can a workflow be engineered to create conception?
• How can inventorship be corrected?
• How did the EPO reject the DABUS applications?
• How did the Federal Circuit resolve Thaler v. Vidal?
• How did the UK Supreme Court split the DABUS problem?
• How does copyright mirror the inventorship debate?
• How does the 2025 guidance characterize AI?
• How does the EPO approach AI inventions in practice?
• How should an invention-assignment agreement handle AI work?
• How should an invention-disclosure form change for AI work?
• How should the choice about AI inventorship be made?
• Is inventorship determined per patent or per claim?
• Is inventorship the only gate an AI-assisted invention must clear?
• Is "name anyone; we can fix it later" a viable strategy?
• Is "who invented it?" a riddle about machine consciousness?
• Must an inventor perform the reduction to practice herself?
• What are the Pannu factors?
• What are the policy stakes underneath the doctrine?
• What arguments favor recognizing AI inventorship?
• What arguments oppose it?
• What assumption did the patent system run on for two centuries?
• What did the AIA change about correcting inventorship?
• What did the USPTO's February 2024 guidance establish?
• What does Section 112 require of an AI-assisted patent?
• What does the directed campaign look like on the strong end?
• What door did the EPO point at?
• What happened in Australia?
• What is the 2025 guidance's touchstone?
• What is the bonus of a directed, iterative workflow?
• What is the debate over the 2025 approach?
• What is the duty of candor and what does breaching it cost?
• What is the hard middle case?
• What is the legal definition of conception?
• What is the most consequential sentence in Thaler v. Vidal?
• What is the pattern across all the worked examples?
• What is the practical effect of treating AI as lab equipment?
• What is the single most important practice for AI-driven R&D?
• What is the structural lesson across patent and copyright?
• What is the subtle chain-of-title trap of an omitted inventor?
• What is the through-line of the entire field?
• What is the through-line of the global DABUS results?
• What middle paths did Germany and Switzerland take?
• What perverse incentive did the 2024 framework create?
• What practical guidance is stable even where the doctrine is not?
• What priority trap does the 2025 guidance add for foreign filings?
• What two cautions keep the 2025 guidance from being a free pass?
• What two rules govern the human-side inventorship analysis?
• What was the DABUS litigation campaign?
• What were the five principles of the 2024 guidance?
• When does secrecy fail as a strategy?
• Why couldn't Thaler claim ownership by owning the machine?
• Why did South Africa grant a DABUS patent?
• Why did the 2025 guidance abandon Pannu?
• Why does AI strain the disclosure requirements?
• Why does documentation matter more under the 2025 guidance, not less?
• Why does inventorship matter financially, not just philosophically?
• Why does ownership have to run through people?
• Why does the provenance of AI inputs matter to a patent program?
• Why does the UK ownership holding land so hard?
• Why do trade secrets sidestep the inventorship question entirely?
• Why is a research goal not conception?
• Why is international harmonization on AI inventorship hard?
• Why is prompting a model and patenting the output doctrinally perilous?
• Why is the one-line prompt the weak case?