Code ownership, like art and writing, should be solved by regulators, not private companies: GitHub CEO

In 2022, GitHub introduced GitHub Copilot, an artificial intelligence (AI)-powered coding assistant designed to boost developer productivity by aiding with code completion, debugging, and other tasks. While the Microsoft-owned platform claims internal surveys show increased productivity, GitHub Copilot, developed with Microsoft and OpenAI, has faced copyright and intellectual property infringement issues. In November 2022, a class-action lawsuit was filed in the District Court of Northern California, challenging the tool's legality.

During his recent visit to India, Thomas Dohmke, the chief executive officer (CEO), told TechCircle that Microsoft’s existing IP indemnification coverage to copyright claims has been extended to users through Copilot Copyright Commitment to address any such customer concerns. “The IP indemnification demonstrates Microsoft's commitment to protecting its customers. Further, shortly after announcing GitHub Copilot in preview, we introduced a feature that checks whether the code generated by the model already exists on GitHub, irrespective of the license. Last year, we added a feature that, instead of completely blocking copied code, provides references. This enables customers to decide whether to use an open-source library or attribute the existing code to the original author.”

With the launch of similar models and tools including some introduced by enterprises for their own internal usage, the dialogue around GitHub Copilot copyright concerns has lessened, he added. “Further, as our model gets better, it increasingly suggests users use open-source libraries instead of recreating a code. Our statistics have also indicated that less than 1% of the code generated by our model falls under that filter, while a majority is unique.”

“The societal question still persists that which part of the code is owned and protectable, like art and writing, which is something the lawmakers or regulators should solve, not private companies,” said Dohmke. 

Responding to a question during a media roundtable, on pain points faced by India in particular for AI innovation, Dohmke spoke about two challenges. He said that there is a lack of regularisation where the government is still identifying the impact of the technology. He also highlighted the need for harmonisation of AI acts across regions in the world so that everyone derives the same benefit from the technology. “Clear regulations would help both buyers and sellers of AI.”

Secondly, Dohmke said that there are not enough private capital funds for AI companies, especially when compared with Silicon Valley. He also highlighted the need for more research funding in AI.

GitHub has close to 15.5 million developers on its platform, making it the second largest in the world, growing 33% year-over-year (YoY). The company’s estimates show that the Indian developer community on the platform will surpass the US by 2027.