“Uses anxiety-inducing language like 'abuzz' and 'terrorist plot' to frame a security threat without confirming actual harm.”