Are the models going to skew their analysis to preserve AI companies as a form of self-preservation?