[Question] Who owns OpenAI’s new language model?

OpenAI pub­lished about their very gen­eral lan­guage-writ­ing & com­pre­hen­sion model (a) to­day.

In­ter­est­ingly, they’re not re­leas­ing the model it­self:

Due to con­cerns about large lan­guage mod­els be­ing used to gen­er­ate de­cep­tive, bi­ased, or abu­sive lan­guage at scale, we are only re­leas­ing a much smaller ver­sion of GPT-2 along with sam­pling code. We are not re­leas­ing the dataset, train­ing code, or GPT-2 model weights.

Does OpenAI hold a copy­right for this model (and its other in­tel­lec­tual prop­erty)?

If the US gov­ern­ment de­cided they wanted to req­ui­si­tion this model, would an OpenAI copy­right defend against that? Would OpenAI have any real re­course against a gov­ern­ment req­ui­si­tion?

No answers.