Criteria for Resolution:
The text of the conversations from the AI box escape experiment must be publicly revealed by either Eliezer Yudkowsky or his counterparties in the experiment. If it is revealed by a counterparty, Yudkowsky must confirm the authenticity of the text. There were multiple instances of the experiment, so revealing one instance will qualify. Revealing only that part of the text which contains the arguments directly leading to the escape will also qualify.
The experiments were performed under conditions of confidentiality. I would not expect Yudkowsky to unilaterally break such confidentiality. Hypothetically, if someone were to want to see the text of one of these conversations posted publicly, I think they'd need to get both parties in one of the conversations to agree to do so.
Has anyone asked @EliezerYudkowsky for his cheerful price (https://www.lesswrong.com/posts/MzKKi7niyEqkBPnyu/your-cheerful-price) for releasing one of the conversations, conditional on the counterparty for that conversation having likewise agreed to do so? Has anyone asked the counterparties?