RSAC 2026 CONFERENCE – San Francisco – External, internal, and operational pressures to deploy AI to unlock its promise of increased speed and efficiency has left enterprise cybersecurity professionals in a tough spot — finding they need to enable innovation, while trying to foresee the risks it might introduce This article explores socs related ai. . Cybersecurity attacks pose a big risk to both of their business environments.
Ankit Gupta is in charge of a Fortune 500 food manufacturing company, and Shilpi Mittal is in charge of protecting a financial company. Both of them agreed to a six-month trial period to see how AI could help them in their security operations centers (SOCs).
Related: AI Dominates RSAC Innovation Sandbox ## AI in the Fortune 500 Food Manufacturing SOC Mittal says she had success using a large language model (LLM) as a "read-only triage assistant" in her food manufacturing company's (SOC) case workflow. She talked about this in an interview with ZeroOwl.
Related: Why Stryker's outage is a wake-up call for disaster recovery AI could also make existing playbooks better, which Gupta has said are "deterministic and rigid, working well only when patterns are predictable." Gupta says, "SOC reality is messy—alerts come with missing fields, inconsistent identifiers, and unclear signals." He goes on to say, "AI wrongly removed users from the system."
He came to the conclusion that AI could help in the SOC, but that people would always have to make the final decisions. He says that LLMs are especially good at summarizing important information, linking context, and making structured narratives from inputs from different security tools. Instead of replacing security analysts or taking full control of alert management, they help by connecting the dots.
On the plus side, Gupta did see a clear drop in analyst fatigue during the pilot period in his financial company. Gupta says, "The biggest change was cutting down on context switching and writing the same thing over and over." "Analysts used to spend 10 to 15 hours a week making documents and getting information for the business.
This work has now been given to AI, and the results are great."" The trial runs are well-timed because leaders in almost every field are under pressure to use AI tools.
