Technical industry may be reeling from a wave of layoffs, a dramatic crypto crash, and the ongoing turmoil on Twitter, but despite these clouds, some investors and entrepreneurs are already eyeing a new boom — one based on artificial intelligence that can create coherent text, engaging images, and functional computer code. But this new frontier has its own cloud.
A class-action lawsuit filed in California federal court this month targets GitHub Copilot, a powerful tool that automatically writes working code when a programmer starts typing. The coder behind the lawsuit claims that GitHub is infringing copyright because it fails to provide attribution when Copilot reproduces open source code covered by a license that requires it.
The lawsuit is in its early stages and its prospects are unclear because the underlying technology is new and has not undergone much legal scrutiny. But legal experts say it could be related to a broader trend of generative AI tools. The AI programs that generate paintings, photos, and illustrations from a prompt, as well as text for marketing copy, are all built using algorithms trained on previous work created by humans.
Visual artists were the first to question the legality and ethics of artificial intelligence, which includes existing work. Some people who make a living from their visual creativity are upset that AI art tools trained on their work can create new images in the same style. The Recording Industry Association of America, a music industry group, said artificial intelligence-assisted creation and remixing of music could be a new area of copyright concerns.
“This whole arc that we’re seeing right now — this generative AI space — what does it mean for these new products that they absorb the work of these creators?” says Matthew Butterick, the designer, programmer and lawyer who filed the lawsuit against GitHub.
Copilot is a powerful example of the creative and commercial potential of generative AI technology. The tool was created by GitHub, a subsidiary of Microsoft that hosts code for hundreds of millions of software projects. GitHub did this by training an algorithm designed to generate code from running AI OpenAI on a vast collection of code it stores, creating a system that can preemptively terminate large chunks of code after a programmer makes a few keystrokes. A recent study by GitHub shows that coders can complete some tasks in less than half the time it usually takes when using Copilot as a helper.
But how some coders were quick to notice, Copilot will occasionally play recognizable snippets of code taken from millions of lines in public code repositories. The lawsuit, filed by Butterick and others, accuses Microsoft, GitHub and OpenAI of copyright infringement because the code does not include attribution, which is required under the open source licenses that cover the code.
Programmers, of course, have always studied, learned, and copied each other’s code. But not everyone is convinced that AI will do the same, especially if AI can produce tons of valuable code on its own without complying with the licensing requirements of the source material. “As a technologist, I’m a big fan of artificial intelligence,” Butterick says. “I’m looking forward to all the possibilities of these tools. But they must be fair to everyone.”
