Schneier recaps the [un]prompted AI security conference and his three-part split of AI security
Bruce Schneier reports on attending the [un]prompted AI security conference in San Francisco, which he describes as three days of dense talks on many aspects of AI security. He divides the field into security of AI, security with AI, and security from AI, and says the event covered all three. He argues that researchers should know which of the three they work on and that the community needs to cover all of them. His own talk, on integrous AI, is not yet online, and he plans to post it later. The post names no vulnerability, product, incident or threat actor, so it carries no direct operational risk. The conference was a practitioner event held October 27-29, 2026, in San Francisco (infosec-conferences.com).
Linha do tempo
-
Schneier recaps the [un]prompted AI security conference and his three-part split of AI security
ler a análise desta fonte ocultar
Bruce Schneier reports on attending the [un]prompted AI security conference in San Francisco, which he describes as three days of dense talks on many aspects of AI security. He divides the field into security of AI, security with AI, and security from AI, and says the event covered all three. He argues that researchers should know which of the three they work on and that the community needs to cover all of them. His own talk, on integrous AI, is not yet online, and he plans to post it later. The post names no vulnerability, product, incident or threat actor, so it carries no direct operational risk. The conference was a practitioner event held October 27-29, 2026, in San Francisco (infosec-conferences.com).