"There’s been a lot of talk lately about the role of AI and copyright, with
much of it focused on fretting by various copyright maximalists about how
things created by AI need more copyright or how AI systems are violating the
copyright of artists, both of which seem to be fairly questionable claims at
But, copyright law professor Brian Frye recently participated in a Copyright
Office “listening session” regarding copyright and AI, and he suggested an
entirely different way that everyone (including the Copyright Office) should be
thinking about. As he notes, the questions everyone seems to be fretting about
appear to be easily answered:
I think we are asking the wrong questions about AI and copyright. Everyone
is asking whether copyright protects AI-generated works and whether training
an AI algorithm infringes copyright. The obvious answer is no and no.
Copyright only protects works created by people. AI doesn’t even create
works, it generates content, which we consumers interpret as works. Roland
Barthes predicted the death of the author, and AI has written the author’s
Likewise, training an AI algorithm doesn’t and shouldn’t infringe copyright.
AI algorithms don’t copy works, they merely catalog rhetorical conventions
and then deploy them to create conventional content.
Instead, he notes, everyone is missing the much bigger picture in that we could
be (and arguably should be) using AI to tell us which other works (of the ones
created by humans) even have enough creativity to deserve copyright protection
in the first place:
We should be asking what AI can tell us about what copyright should protect
and why. Copyright can only protect “creative” works. But courts and the
Copyright Office have struggled to define “creativity.” Maybe AI can help?"
Share and enjoy,
*** Xanni ***
Chief Scientist, Xanadu
Partner, Glass Wings
Manager, Serious Cybernetics