Image: © LT/Stock.adobe.com
The Théâtre D’opéra Spatial is being denied copyright protection due to the amount of AI-generated content it contains.
An AI-generated piece of art that caused controversy last year for taking a top prize has been denied copyright protection in the US.
The artwork – by Jason Allen – took first prize in the Colorado State Fair’s fine art competition last year. This led to criticism when it was revealed that the piece was created using Midjourney, a popular text-to-image AI generator.
These image generators take a person’s typed prompts and convert them into an image – other text-to-image AI generators include Google’s Imagen AI and OpenAI’s DALL-E 2.
In a filing last week, the US Copyright Review Board said Allen’s artwork titled Théâtre D’opéra Spatial was denied copyright protection due to the amount of AI-generated material it contains. The copyright request was denied last year, but Allen has asked the board to reconsider the request twice.
The board said Allen first attempted to submit the work for protection on 21 September last year, but he did not disclose that the artwork was made using an AI system. However, the Copyright Office claims it was already aware that the art contained AI-generated material due to the “national attention” it garnered after it won the award.
Allen later told the Copyright Office that he input “numerous revisions and text prompts at least 624 times to arrive at the initial version of the image”. He also said that he used Adobe Photoshop and Gigapixel AI to remove flaws, create new visual content and upscale the image.
The Copyright Office asked Allen to disclose the parts of the image generated by Midjourney to be eligible for copyright protection, but Allen declined. Last year, Allen told a discord server that fine-tuning the text prompt and editing what Midjourney produced constituted “at least 10pc” of the work.
“Because Mr Allen is unwilling to disclaim the AI-generated material, the work cannot be registered as submitted,” the board said in its filing.
AI vs copyright
The decision will likely mark another blow for those attempting to copyright AI-generated artwork. Last month, a US court ruled that artwork generated by AI cannot be copyrighted. US district court judge Beryl Howell said copyright has never been granted to work that was “absent any guiding human hand” and that human beings are an “essential part of a valid copyright claim”.
The debate on AI and copyright has been intensifying ever since generative AI tools such as ChatGPT became popular.
In July, thousands of authors including Margaret Atwood and Jodi Picoult signed a letter calling on the likes of OpenAI, Alphabet and Meta to stop using their work to train AI models without “consent, credit or compensation”.
This debate also comes at a time when the Writers Guild of America continues to spearhead the Hollywood strike, with complaints that include the use of AI tech without fair compensation for screen writers and actors.
The US Copyright Office has asked the public for its views on this issue. Public comments are welcome on the office’s website until 18 October.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.
