Amid seemingly daily advances in the sophistication of artificial intelligence (AI) platforms such as ChatGPT, DALL-E and Midjourney, the U.S. Copyright Office last week published a policy statement clarifying the standards the office has followed—and will continue to follow—in determining whether an expressive work created with the involvement of AI is copyrightable. The policy statement provides some clear guidelines for users of some of the more popular AI platforms, while still making clear that the office’s analysis will largely depend on the particular facts of each application.
The policy statement reaffirms the core principle that “copyright can protect only material that is the product of human creativity.” In order for a work to be subject to registration, the “work must be created by a human being,” and the office “will not register works produced by a machine or mere mechanical process that operates randomly or automatically without any creative output or intervention from a human author.” Thus, when the office considers whether a work created with the involvement of AI may be copyrighted, it will ask “whether the ‘work’ is basically one of human authorship, with the computer [or other device] merely being an assisting instrument, or whether the traditional elements of authorship in the work … were actually conceived and executed not by man but by a machine.”
The emphasis on human authorship sets the standard for what sorts of expressive work will not be subject to registration. For example, works created as the result of a human feeding a text prompt into an AI platform—as is the case for visual works from DALL-E or written works from ChatGPT—will not be deemed copyrightable. In those scenarios, the “prompts function more like instructions to a commissioned artist—they identify what the prompter wishes to have depicted, but the machine determines how those instructions are implemented in its output.” On the other hand, the office indicated that works that use AI-generated material as a base, but then layer human creativity on top of that base—such as by “select[ing] or arrang[ing] AI-generated material in a sufficiently creative way” or “modify[ing] material originally generated by AI technology”—may meet the standard for copyright protection. The latter sort of applications will involve a more fact-intensive analysis. And, importantly, “copyright will only protect the human-authored aspects of the work.”
Given the above parameters, the policy statement reminds applicants that they “have a duty to disclose the inclusion of AI-generated content in a work submitted for registration and to provide a brief explanation of the human author’s contributions to the work.” It is incumbent upon applicants to identify the extent to which AI was used to create a work in order to ensure that the office only approves human-authored content.
Summary prepared by Wook Hwang and Edward Delman