The fact that training data is unregulated is absurd. The leading voices on the data side are making a case for some form of regulation, largely because they want it to be forward-looking, unapologetic for the prior behavior of theft culture. What this doesn’t account for is clear and existing risk of consumer harm. The genie is out of the bottle, but that’s no excuse for lack of action. 

This is a safety issue. Our children are excessively digitally native, and often behave with a naivete around their digital presence. As we look at social media, we all wish we had a better handle on its clear effects. It is inarguably harmful. In hindsight, we know it should have been better regulated. We must not make the same mistake with AI.