Listen to this story (00:24 - 07:15):

The Story: Venture capital is all-in on artificial intelligence. But if a few copyright lawsuits don’t go the way VC would like them to, the industry could be holding a very pricey AI bag.

Generative AI models rely on absurd amounts of data points to train their chatbots. For many models, the data on which they train is data protected by U.S. copyright laws. Copyright holders like Universal Music Group have filed lawsuits against AI companies for training their models on said protected data. And if these lawsuits are decided in favor of copyright holders, the cost structures of AI companies would be seriously disrupted and VCs could lose billions on their investments.

But VCs who invest in generative AI companies do not appear to be concerned. In an interview with Axios’s Dan Primack, one VC, who only spoke on background, stated that, quote, “AI training in most cases is not analogous to copying or plagiarism…These models for the most part do not plagiarize or copy data they were trained on, and they are not databases with lookup tables. As a result, they are not equivalent to the precedent set by Napster and others in the music piracy era.”

Another unnamed VC who is not particularly concerned responded to Primack saying that he suspects the firms who are heavily invested in AI will have “researched this with their in-house lawyers before writing a billion dollar (or multi-billion dollar) check.”

The Expert Take: Lily Lyman, general partner at Underscore VC, believes a compromise will eventually be made between AI companies and copyright owners. “I think we’re gonna see a world where [copyrighted materials and AI models] can actually coexist,” says Lyman.

Rob Biederman, managing partner at Asymmetric Capital, joins other VCs in voicing that he is not terribly concerned about these lawsuits. He believes enforcing the prohibition of AI models from training on copyrighted materials would be “very very difficult.”

While the legality of training chatbots with copyrighted data remains uncertain, investors across venture capital seem confident AI companies will escape unscathed from infringement accusations.