Why is it that all of a sudden with generative AI, what makes copyright so much more complex in this new AI environment?
AI has been around for over 70 years, and we’ve been interacting with AI on a daily basis for probably the last 15 or even 20 years.
But AI has evolved from being essentially about predicting outcomes to creating content.
And to do that, it was trained on vast amounts of existing works. Things like books, images, articles, blogs, code, and most of the works that it was trained on was protected by copyright.
Now the outputs that the AI generates can also compete with the very works that the AI was trained on. It could also reproduce parts of those works. So that’s one of the reasons why copyright has become so central to the AI conversation.
And that’s a fundamental shift that introduces new and important liabilities.