A huge near-term risk of AI is state actors using them as pseudo-spies. It is easily possible to deploy a swarm of manipulative chatbots as “spies” to seduce people with access to important information and systems to give up those facts.

→ View original post on X — @alexandr_wang,