"""That could, for instance, be data showing that women in certain parts of the U.K. are more liable to read Covid misinformation, or that certain teens are “hyper-exposed” to self-harming content. Ofcom would then tell the social media firm to tweak its algorithms to change those statistics, or be punished."""<p>"Lies, damned lies, and statistics" came to mind. I also strongly suspect that the internal research would get outsourced or otherwise far more difficult to discover.