A US$9 billion lawsuit is filed against Microsoft and GitHub over infringing copyrights

A proposed class action complaint names Microsoft and GitHub, and its business partner OpenAI, alleging that their development of the AI-powered coding assistance GitHub Copilot depends on "software theft on an unprecedented scale." The case is still in its early stages, but it has the potential to have a significant impact.

It will impact the wider field of artificial intelligence, where businesses are generating fortunes by teaching software using data that is copyright-protected. Copilot, which Microsoft-owned GitHub launched in June 2021, is trained on web-scraped public code repositories, many of which are made available under licenses that oblige anyone using the code to give credit to its authors. Copilot has been found to repeat lengthy passages of licensed code without giving credit, which has led to this complaint accusing the corporations of flagrant copyright law violations. A set of 11 open-source licenses that require attribution of the author's name and copyright.

Programmer and attorney Matthew Butterick, who filed the complaint with the aid of the San Francisco-based Joseph Saveri Law Firm, stated in a press release that "We are chal­leng­ing the legal­ity of GitHub Copi­lot." "This is just the beginning of a long journey. According to our knowledge, this is the first class-action lawsuit in the US to contest how AI systems are trained and produced. It won't be the final. AI software is not excluded from the law. These systems' creators and operators must continue to be held accountable.

Microsoft and OpenAI are not the only companies using web content that is protected by copyright to train AI systems. The open-source software Stable Diffusion is only one example of a text-to-image AI that was developed in precisely the same manner. The companies behind these programs believe that the fair use theory in the US covers their use of this data. However, according to legal experts, this is not yet settled law, and future litigation, such as Butterick's class action case, might completely alter the ill-defined status quo.

GitHub responded to a request for comment by saying: "We've been committed to responsibly innovating with Copilot from the outset, and will continue to evolve the product to best serve developers throughout the world."