I flatly discount the possibility because OpenAI can't produce evidence of a breach. At best, they'd rather hide the truth than admit a compromise. At worst the

I flatly discount the possibility because OpenAI can't produce evidence of a breach. At best, they'd rather hide the truth than admit a compromise. At worst they show incompetence that they couldn't detect such a breach. Not a good look either way.