Events
,
RSAC Conference
,
RSAC Conference Featured
Vice President Singer on How Custom AI Models Improve Security, Cost and Control
General-purpose artificial intelligence models were not built for cybersecurity, and the gap is showing. Organizations using large off-the-shelf models face three compounding challenges: limited customizability, data privacy risks and the steep cost of running oversized models on narrow, specific tasks, said Yaron Singer, vice president of AI and security at Foundation AI, Cisco.
The answer lies not in scaling models up, but in building them right. Smaller models trained exclusively on cybersecurity data are inherently domain-specific, fine-tunable for precise tasks and deployable without routing sensitive data to external providers. That combination of control, privacy and cost efficiency is what makes them better suited to security operations than general-purpose alternatives.
“If by bringing in a model and technology around it, one can basically help with unlocking some of the analyst capabilities, because they can do things that normally would take them a month, they can now do them in 10 minutes – that’s groundbreaking,” Singer said.
In this video interview with Information Security Media Group at RSAC Conference 2026, Singer also discussed:
- How Foundation AI’s expanding portfolio of base, reasoning and enterprise search models enables the development of agentic security systems;
- Role of AI agents in SOC environments and workflow automation;
- Why production AI systems require continuous maintenance as models age and APIs evolve.
Singer leads AI and engineering for Cisco Security’s Foundation AI team, advancing secure AI technologies across Cisco and Splunk. Previously, he co-founded Robust Intelligence, an AI security pioneer acquired by Cisco, and built deep expertise spanning AI, machine learning, cybersecurity and applied research.
By clicking here, you agree to receive contact from Cisco with offers, promotions, and the latest news regarding its products and services. You can unsubscribe at any time. Click here to learn more.
