If trust is our currency with teams and clients, can we afford to delegate all our communication to algorithms?

Should we be worried that, when stress tested, AI agents have a propensity to resort to blackmail, espionage and worse?