Salesforce CISO on Earning AI Autonomy in Agentic Security Operations
Iain Mulholland, EVP and CISO at Salesforce, spoke about agentic security operations at Dreamforce. Agentic security refers to using artificial intelligence (AI) agents to automate security tasks.
Mulholland emphasized the importance of understanding data before deploying AI agents. He also stressed that autonomy should be earned through data and not come at the expense of guardrails, which are measures put in place to prevent AI agents from making decisions that could harm the organization or its customers.
Salesforce has developed several tools to support agentic security operations, including the Trusted Enterprise AI Harness and Salesforce Guardian. The company has also announced a kill switch feature that allows administrators to pause AI agents quickly if they detect any issues.