The AI-dominated RSAC 2026 conference saw the launch of new agentic security tools from vendors including Saviynt, CrowdStrike and Check Point.
Focus On AI Agents
The ubiquitous, nearly exclusive focus on AI and agents at RSAC 2026 made it clear that the cybersecurity industry is reshaping itself around these still-developing technologies. But while the taglines from vendors about their new offerings may sound similar, many of the new products unveiled at RSAC in San Francisco this week in fact covered a variety of segments within the AI and agentic space.
[Related: 5 Cybersecurity Companies Making Big AI Moves At RSAC 2026]
CRN was on hand to survey the new offerings on the show floor and meet with top executives from top industry vendors this week. We covered major product launches on Monday and Tuesday this week, and we’ve now collected 10 more product announcements that have been on our radar from key vendors including Saviynt, CrowdStrike and Check Point.
The product launches underscore the industry’s growing focus on runtime enforcement for AI, in order to prevent risks from AI and agents in real time, while also bringing an emphasis on the central importance of securing identity when it comes to agentic. Additional product launches aim to enable the agentic-enabled SOC (Security Operations Center), with the goal of automating a far greater portion of the basic tasks of the SOC analyst. Other announcements have targeted security for the massive quantities of AI-generated software code currently being created.
Here’s what you need to know on 10 cool new AI and agentic tools announced this week at RSAC 2026.
Saviynt Identity Security for AI
At RSAC 2026, Saviynt debuted its new Identity Security for AI platform, bringing together a number of identity security capabilities aimed at protecting the usage of AI agents. The launch is Saviynt’s “most significant release for AI agents and LLMs” to date, Saviynt CEO Sachin Nayyar said in an interview with CRN. “We think you need all the pieces of identity management—core identity management, posture management, privileged access management, vaulting, enforcement—everything running together at AI speed,” Nayyar said. “We built all the components ourselves. So everything is connected. We are able to put everything together at runtime for enforcement.” Saviynt is well ahead of the competition in enforcement for AI agents due to offering fine-grained access capabilities, he said. By delivering all of these capabilities together, Saviynt believes it is providing the “most comprehensive identity security capability in this space,” Nayyar said.
CrowdStrike: Charlotte AI AgentWorks Ecosystem
At RSAC 2026 on Wednesday, CrowdStrike unveiled a new collaboration around Charlotte AI AgentWorks, which is the company’s no-code platform for building, testing, deploying and orchestrating security agents that debuted in September. The new collaboration, the Charlotte AI AgentWorks Ecosystem, features a number of major launch partners including Accenture and Deloitte as well as Anthropic, OpenAI, AWS, Nvidia, Salesforce and Kroll. The collaboration is aimed at accelerating the generation of secure agents using Charlotte AI AgentWorks, while also ultimately enabling the creation of agentic security businesses around CrowdStrike’s Falcon platform. The move by CrowdStrike and its collaborators is poised to help deliver “greater customization to the agentic SOC, helping clients build tailored agents they can trust,” said Rex Thexton, CTO at Accenture Cybersecurity, in a quote included in a news release.
Darktrace Adaptive Human Defense
Darktrace announced at RSAC 2026 that it has released its new Adaptive Human Defense offering to offer an AI-era alternative to traditional security awareness training. Adaptive Human Defense offers adaptive coaching for employees in real time, leveraging behavioral AI to educate users during the work day while also “notifying them of risky behavior and providing short, relevant coaching on those risks before bad habits form,” the company said in a news release. The offering then utilizes the results of the “micro-coaching” sessions to adjust safeguards within the employee’s email inbox, Darktrace said.
Check Point AI Defense Plane
At RSAC 2026, Check Point Software Technologies debuted its new Check Point AI Defense Plane offering with an aim toward delivering enhanced security for agentic-enabled enterprises. AI Defense Plane delivers key modules including security for workforce AI, which offers visibility, governance and runtime protection for AI usage by employees. The offering also provides security for AI applications and agents—via discovery, posture and controls at runtime—as well as AI red teaming capabilities.
Absolute Security: Agentic Cyber Resilience
At RSAC 2026, Absolute Security unveiled its new approach to integrate agentic AI capabilities into its Cyber Resilience Platform. Through the addition of agentic AI going forward, the company aims to “equip security leaders and Chief Financial Officers (CFOs) to adapt and remain ahead of future business and security challenges,” the company said in a news release. The announcement came as Absolute Security also debuted its new Rehydrate Ready capability, which allows supported endpoint devices to be placed into a state that is capable of recovery without upfront cost or need for licensing, Absolute Security said.
Databricks Lakewatch
Databricks announced at RSAC 2026 that it is expanding into the cybersecurity space, with the debut of a new agentic SIEM product that taps into the data management and AI capabilities of the company’s core Data Intelligence Platform to provide threat detection and investigation capabilities. The company said that its new Lakewatch offering is designed to help organizations defend against increasingly sophisticated attackers using AI and agents that reduce the time security teams have to detect and respond to attacks. Key capabilities include agentic triage and investigation as well as automated security intelligence, Databricks said.
Cyderes Meridian
Cyderes announced at RSAC 2026 that it has enhanced its Cyderes Managed Detection and Response offering with the introduction of a new entity fabric to boost threat investigations. The new AI-powered entity fabric, Meridian, delivers a “continuously reconciled understanding of how identities, assets, access and exposures connect across the enterprise,” the company said in a news release. Meridian ultimately will reduce cyber risk by serving as a “foundational intelligence layer” driving Cyderes’ MDR offering, enabling investigation of threats with complete context about IT environments, the company said.
Varonis Atlas
At RSAC 2026, Varonis unveiled its new “end-to-end” AI security platform with the launch of Varonis Atlas. The offering spans discovery, assessment and remediation for agents and LLMs, the company said. Key capabilities include continuous discovery of AI assets, including unsanctioned shadow AI, as well as AI security posture management (AI-SPM) and AI penetration testing. Other functionality covers remediation with AI runtime guardrails as well as AI compliance and governance, Varonis said. In terms of detection, Atlas offers AI detection and response (AIDR) along with AI activity monitoring, the company said.
Sumo Logic announced at RSAC 2026 that it has expanded its Dojo AI Agents offerings to include additional capabilities for the AI-enabled SOC (Security Operations Center) analyst. Key capabilities now in preview include SOC Analyst Agent, which enables reduced MTTR (mean time to remediation) with functionality spanning from “automated to human-led investigations, to context-aware response actions and recommendations,” the company said in a news release. Other capabilities now in preview include the Sumo Logic MCP Server, while functionality including Query Agent and Knowledge Agent are now generally available, Sumo Logic said.
Apiiro announced at RSAC 2026 that it has added a key new capability to its Guardian Agent offering that delivers automated “architecture-aware” threat models. The new capability, AI Threat Modeling, provides the ability for security teams to keep up with the pace of code generation in the AI era, when a large portion of new software is being created by LLM-powered tools and agents. Key functionality in AI Threat Modeling includes rapid modeling of threats for every artifact, analysis that is based on actual software architecture and prevention of security risk prior to the existence of code.
