Robots to be sued for copyright breach
COMPUTER programmers plan to sue Microsoft over breach of their copyright in the code they've written - by a so-called "artificial intelligence" system. Like systems that could replace stock photos, Microsoft's GitHub Copilot is a "machine learning" system - it is "trained" on the creative work of humans.
A robot writing computer code, generated from that prompt by Dall-e-2
"We are challenging the legality of GitHub Copilot," programmer and lawyer Matthew Butterick told The Verge. The lawsuit, filed on 4 November through Joseph Saveri Law Firm in San Francisco, "is the first step in what will be a long journey. As far as we know, this is the first class-action case in the US challenging the training and output of AI systems... Those who create and operate these systems must remain accountable."
The case will also be a test of "open source" licences. Many programmers choose to publish code on condition that they are credited. Though there is no moral right of attribution in US law*, reproducing code without a credit would be a breach of the economic rights that those licences deal with. And, indeed, James Vincent writes in The Verge, "Copilot has been found to regurgitate long sections of licensed code without providing credit."
Copyright conundrum
If the case gets very far it will also produce an interesting take on US "fair use" doctrine. The use of code by GitHub Copilot is probably "transformative" - a criterion that US courts have developed in declaring uses to be fair. Much will hinge on the amounts and substantiality of the portions used and the effect on the market for the code - as well as the apparently flagrant breach of open source licence terms.
A ruling on this would have seem to have direct effect on the legality in the US of, for example, images generated by "learning" from stock image libraries - such as the one above. In UK law, copyright in computer code is probably more distinct from that in images.

