US Judge Denies Copyright For AI-Generated Artwork

robot Image credit: Pexels

US judge denies copyright to computer scientist for AI-generated artwork, calls human authorship ‘bedrock requirement’

A Washington, DC judge has ruled that artwork created by an artifical intelligence (AI) without human input cannot be copyrighted under US law.

The Friday ruling by US District Court Judge Beryl A. Howell confirmed an earlier US Copyright Office decision to deny a copyright for an artwork to Stephen Thaler, who had applied for it on behalf of his AI, the Creativity Machine.

The AI-generated artwork, called “A Recent Entrance to Paradise”, was created without human interaction by “a computer algorithm running on a machine”, according to Thaler’s copyright application, which listed the Creativity Machine as the author of the work and Thaler as “the owner of the Creativity Machine”.

Thaler has also applied for patents in several countries for inventions created by his DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) AI, with limited success.

AI Generated image copyright Image credit Stephen Thaler Creativity Machine
‘A Recent Entrance to Paradise’. Image credit: Stephen Thaler/Creativity Machine

‘Bedrock requirement’

Judge Howell sided with the Copyright Office, saying human authorship was a “bedrock requirement of copyright”.

She acknowledged that “we are approaching new frontiers in copyright” as artists begin using AI as a tool, which would raise “challenging questions” for copyright law.

But she said the Thaler case was “not nearly so complex”.

Thaler’s attorney, Ryan Abbott, said on Monday he and his client strongly disagreed with the decision and planned to appeal, while the Copyright Office said in a statement that it “believes the court reached the correct result”.

Data scraping

The US Copyright Office has also denied a copyright to an artist who used the generative AI Midjourney to create a work, saying the system was part of their creative process.

Meanwhile several lawsuits have alleged that the way generative AI systems are trained using vast amounts of existing material amounts to a large-scale piracy.

Those lawsuits include one filed by Sarah Silverman and two other authors against OpenAI and Meta earlier this year over their data-scraping practices.

Programmer and lawyer Matthew Butterick last November filed a similar class-action lawsuit against Microsoft, its subsidiary GitHub and Microsoft-backed ChatGPT creator OpenAI over OpenAI Codex, a programming tool that powers GitHub’s coding assistance feature GitHub Copilot.