RSAC 2026: Agentic AI Dominates the World's Largest Cybersecurity Conference

The RSA Conference 2026, the world's largest cybersecurity event, is underway in San Francisco from March 23 to 26, bringing together over 700 speakers, more than 570 sessions, and upwards of 600 exhibitors. This year, one theme towers above all others: agentic AI and its profound impact on both cyber defense and cyber offense.

The Rise of Agentic AI in Security

The conference has made it clear that AI agents are no longer a future concept but a present reality reshaping enterprise security. From autonomous SOC analysts to self-healing infrastructure, vendors across the industry are racing to embed AI agents into every layer of the security stack.

Varonis CEO Yaki Faitelson delivered a keynote titled "Robots vs. Robots: Stories from the Frontlines of the Agentic Revolution," sharing real-world examples of how AI is reshaping enterprise security across three critical dimensions: defense automation, threat detection, and data governance.

As one industry expert put it at the conference: enterprises now need "agents for every part of your security program, from GRC to IAM to SOC, everywhere."

Key Announcements

AI-Native Security Platforms

• Booz Allen Hamilton launched Vellox, a suite of five AI-native cybersecurity tools covering malware analysis, detection engineering, adversary emulation, compliance monitoring, and autonomous remediation
• Arctic Wolf announced the Aurora Agentic SOC, combining its Concierge Experience with turnkey agentic AI for automated threat response
• Panther announced general availability of its AI SOC Platform, where AI agents have native access to the data lake, detection engine, and organizational knowledge
• Splunk showcased its unified Agentic SOC approach, combining detection, investigation, and response into a cohesive workflow powered by natural language interaction

Securing AI Agents Themselves

• 1Password announced Unified Access, a new agent security platform enabling organizations to securely deploy AI agents and automated workflows
• Geordie AI was named Most Innovative Startup 2026 in the prestigious Innovation Sandbox contest for its security and governance platform purpose-built for AI agents

Autonomous Offensive Security

• Assail launched Ares, an autonomous red-teaming platform designed to discover, chain, and exploit vulnerabilities across APIs, mobile applications, and web applications

Geordie AI Wins Innovation Sandbox

The Innovation Sandbox contest, one of the most watched events at RSAC, crowned Geordie AI as its 2026 winner. Founded by CEO Henry Comfort, the company offers a platform that gives enterprises real-time visibility into their "agentic footprint" and helps them monitor agent behavior and mitigate risks at scale.

The contest has an impressive track record: over 20 years, its Top 10 finalists have collectively achieved more than 100 acquisitions and received over $50 billion in venture funding. Each 2026 finalist received $5 million in investment support.

Beyond AI: Other Critical Themes

Post-Quantum Cryptography

With quantum computing timelines accelerating, discussions have shifted from theoretical to practical. Organizations are urged to inventory their cryptographic assets and develop migration strategies to quantum-resistant encryption now rather than wait.

Continuous Threat Exposure Management

Security teams are moving away from periodic vulnerability assessments toward continuous visibility into attack paths and exposures, with vendors racing to build unified CTEM platforms.

North Korean AI-Powered Threats

Microsoft revealed at the conference that North Korean state actors are now using AI for identity fabrication and long-term persistence at scale, forging agent identities to infiltrate organizations.

What This Means

RSAC 2026 marks a turning point: the cybersecurity industry has fully embraced AI agents as both its greatest tool and its greatest challenge. Organizations that fail to secure their agentic operations risk creating new attack surfaces even as they automate their defenses. The race is no longer just about detecting threats but about governing an entirely new class of autonomous digital workers.

Source: RSA Conference 2026