Who owns the code?

Ownership of AI-generated code remains a contentious issue under U.S. copyright law. Since the law requires a human author for copyright protection, code entirely created by artificial intelligence cannot be copyrighted or owned. The AI system does not possess authorship, so any code it produces lacks human authorship and, consequently, copyright protection.

When AI contributes significantly to the creative process, the resulting code may not be considered human-authored, leaving it unprotected under copyright law. Only the sections of the code where a human actively contributes meaningful input are eligible for copyright protection, while the AI-generated portions remain unowned.

In a mixed codebase combining human and AI-authored components, ownership is limited to the parts a person genuinely wrote. The AI-generated segments do not confer any ownership rights until a human revises or reworks them. Open-source licensing does not negate copyright protection, so a pure AI output subjected to an open-source license lacks enforceable ownership.

Recent legal rulings and reports have solidified this rule, emphasizing that without a human author, there can be no copyright protection. The court's refusal to review a previous decision maintains the human authorship requirement as the settled U.S. copyright standard, barring machines from recognizing themselves as authors. Only human-created works can secure copyright ownership, leaving AI-generated works without legal protection.