Microsoft executive calls for security considerations to address agentic artificial intelligence

SAN FRANCISCO. Use of agentic artificial intelligence tools must be paired with addressing security needs, according to Vasu Jakkal, corporate vice president for Microsoft's security practice, who spoke here at the RSA conference in an opening keynote.

“AI must start with security,” Jakkal said, starting with security considerations for identity and permissions including what AI agents can do, what access they have and who they should be working with. Jakkal said there are also questions around preparing data “to prevent...