Third party AI training dataset, AKA a treasure trove of copyright infringement that you think you're not responsible for profiting from because someone else did the dirty work.
They’re just laundering liability.
If they use a third party data set then (presumably) they won’t be liable when it obviously contains copyrighted material!
There’s a huge market for data sets that are sold as “copyright free” or “ethical” but no one actually audits the data
I think this is what my "friends" did, built a "Christian" content model off of stolen work and then sold the base to people managing Christian influencers to "skin" their clients voice and personality on top of it.
