Federal grant funds plan to detect "misinformation," "hate speech" on Twitter