"Truth": Unmasking "Coded" Hate Speech Extremists often avoid legal and social accountability by using coded language or memes rather than overt Nazi symbols. The Unmasking Antisemitism Project: Run by researchers at American University, this initiative uses machine learning to scan millions of social media posts.The Result: The AI acts as an early-warning system by identifying hidden hate speech and tracing the digital footprints of extremists before they can radicalize others or commit acts of violence.