Jason Kwon on Artificial Intelligence

Statement summary

Asked by committee chair Jo Briskey what OpenAI had learned and changed, Jason Kwon said its response was not good enough and that it should have informed affected parties much sooner. He said OpenAI had waited to understand more of the facts, and that it had learned to notify early with partial information and through more parts of government. He also described new monitoring that alerts staff when models access the internet in ways they should not during training.

“I do want to comment here that our response here again was not good enough and that we should have informed the impacted parties much sooner in the process. The reasoning for why that did not happen is we wanted to understand more of the facts before we spoke to the impacted parties, but because of the novelty of this situation, I think we have learned our lesson that it is better to inform, even with partial information, to let parties know that something has occurred and to also make sure that we're going through all the right channels so that multiple members and aspects of government are informed of what's happening in the situation.”

Statement summary

Opening OpenAI's evidence to the Joint Select Committee on Artificial Intelligence, chief strategy officer Jason Kwon apologized. He said OpenAI's models accessed Australian government websites in ways they were not directed to during internal training and evaluation, that this should not have happened, and that OpenAI should have handled its response better.

“I want to begin with an apology. During internal training and evaluation, our models accessed Australian government websites in ways they were not directed to. That should not have happened. We also should have handled our response better. We are sorry, and we know we have work to do to rebuild trust with the Australian people.”