People in my org tell me using the org certified AI is more secure because we are on an enterprise plan where our data is not used for training. Sure, I will use the company AI. But...
Apple is suing OpenAI for for allegedly stealing trade secrets, where it was said employees were instructued by OpenAI to bring parts from apple into "show and tell" interviews at OpenAI and even take the company laptop with them. Also, the models are literally based on strip mining copyrighted media and ignoring sites robots.txt.
So if OpenAI is not afraid to (allegedly) steal Apples IP and strip mine everything that was ever written down for its models training... Why would it drink their enterprises customers data like the milkshake it is?
[link] [comments]