Meta announced Friday that it will begin labeling a broader range of AI-generated content beginning May in a bid to reassure the public amid increasing risks of deep fakes in a historic global election year.