It's very likely that OAI models will have even more restrictions. Firstly because now they know what feds will do if you don't tune the safety classifiers towards more false positives and secondly, OAI models were always more restrictive than ANT.