The Copyright Office is Wrong About AI-Assisted Works
summary
Claim: the US Copyright Office rejects copyright in Zarya of the Dawn images generated with MidJourney, arguing the outputs are machine-generated and not protectable. The uploader documents hundreds of prompt iterations producing consistent characters and style, but the Office treats that as randomness. The argument compares this to 19th‑century fights over photographs, predicts eventual reversal but over years, and warns that model weights lack IP protection, vendor indemnities are limited, and outputs can still infringe inputs.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:08Copyright Office’s flawed AI policy
- 2:32Legal precedent on authorship
- 5:00Zarya of the Dawn case
- 8:15AI-assisted works need copyright
- 10:20Decline of letterpress printing
- 13:13AI in software development
- 17:07Photography’s copyright journey
- 23:27Copyrightability in AI-assisted works
- 29:13Future of AI copyright law
- 31:43AI-driven content strategies
- 35:16Q&A
related talks
15mWho is the owner of AI-Generated works?
19mHumans vs. Machines
32mGenerative AI & Intellectual Property Law
50mAI Development and Copyright: Public ≠ Free Use
28mAI – Legal Insights
48mWho Wrote the Code? On Copyright for Code Written with AI
37mThe Shadows That Follow AI Generative Models
32mIntellectual Property Considerations for Developing and Licensing Software