You can use AI to fact-check and filter malicious content. (Which would lead to another problem, which is... who fact-checks the AI?)