Soon everybody will be worried about Chinese LLM's as agent Trojan horses

David Mack warned that Chinese large language models could be used as agent Trojan horses, with Michael Herrmann suggesting they could be trained to trigger on specific keywords or phrases.

Soon everybody will be very worried about Chinese LLM’s as agent Trojan horses David Mack@DavidHHMackSoon everybody will be very worried about Chinese LLM’s as agent Trojan horses6:39 AM · Jul 23, 202646Views202202 Michael Herrmann@m herrmann37mCool idea. Can probably train them to trigger on certain keywords or phrases.707