Is that not an acknowledgment/admission that agents would easily penetrate "secure" systems if explicitly directed to do so? This was roughly 1,000 independent agents, imagine 1 million+ directed to work together. The conclusion being that nothing is secure any longer.
