At the recent RSAC 2026, the cybersecurity industry’s major players and innovative startups converged to showcase a new range of tools designed to enhance the security of AI technologies and agents. The overarching theme of AI has dominated the conference for the past several years, but this year, the focus has expanded significantly. As organizations of all sizes grapple with the urgent challenge of securing AI agents, the conference brought forth a striking array of solutions aimed at addressing these concerns.
The rapid adoption of AI agents for productivity has opened organizations to unprecedented security risks, particularly due to these agents’ potential access to sensitive data and their capability to operate autonomously. One of the pressing threats highlighted at RSAC 2026 is the proliferation of unauthorized “shadow” agents—a phenomenon that has made securing AI technologies a central topic for cybersecurity vendors. These agents, if left unchecked, pose significant security risks, driving industry players to focus their product announcements on this critical issue.
Kicking off on Monday in San Francisco, RSAC 2026 featured a series of major product launches from industry giants such as Palo Alto Networks, CrowdStrike, Cisco Systems, and Wiz—a company owned by Google. A wide array of new AI and security products was unveiled, reflecting the concerted effort across the industry to tackle these mounting challenges. Vendors such as Cyera, HPE, and Tenable announced their latest solutions aimed at securing AI and agents on Tuesday, contributing to the growing conversation around this pressing topic.
The conference attracted a variety of companies, including publicly traded firms like Netskope alongside venture-backed startups such as Torq and Chainguard, all unveiling crucial new AI and security products just before the commencement of RSAC 2026. This mix of established and emerging companies showcased the industry’s collective urgency to address the security implications of AI technologies.
Among the standout products presented was CrowdStrike’s new Falcon Data Security offering. This unified architecture spans identity management, endpoint protection, SaaS applications, cloud solutions, and AI, enabling real-time awareness of sensitive data movement. Additionally, CrowdStrike introduced its next-generation Managed Detection and Response (MDR) solution, Agentic MDR, engineered to utilize AI agents for automating high-friction security workflows at machine speed and enhancing protection during critical transactions.
Cyera also made headlines with the announcement of its Browser Shield tool, designed to provide real-time visibility and discovery for AI technologies operating within browser environments, whether managed or unmanaged. This innovative tool facilitates alert-level blocking to prevent data leakage and unethical AI usage, while also creating a data lineage product that maps agent activities, thereby offering greater insights for security teams.
Huntress followed suit, enhancing its Managed ITDR (Identity Threat Detection and Response) offering to support Google Workspace alongside Microsoft 365, thus enabling MSPs to provide a unified managed service regardless of the productivity platform employed by customers. The service includes monitoring for identity abuse, uncertain login activities, and the tracking of malicious data center utilization.
Meanwhile, HPE showcased groundbreaking firewall technology with the introduction of the Juniper Networking SRX400 series, catering specifically to space-constrained environments while offering robust protection against tampering. The extended capabilities also include enhanced visibility and access management concerning AI system usage.
The innovation extended beyond traditional cybersecurity firms, with Torq launching its Agentic Builder, aimed at enhancing operational productivity by allowing security teams to articulate their desired security outcomes in natural language. This empowers the creation of tailored Torq AI Agents for continuous management of security alerts and crucial operations.
SentinelOne presented its Prompt Security On-Premise offering, which provides host-based AI security capabilities to monitor unsanctioned AI usage and facilitate real-time redaction of sensitive information across numerous applications. This product acts as a specialized firewall for both internal and external AI tools, addressing a critical need for organizations prioritizing data security.
Other remarkable unveils included Palo Alto Networks’ Prisma AIRS 3.0, which aims at extensive discovery and assessment of AI agent activities, and Okta’s Secure Agentic Enterprise Framework, designed to tackle questions surrounding agent discovery, connectivity, and functionality.
The range of product offerings and innovations introduced at RSAC 2026 demonstrates the industry’s commitment to navigating the complexities of AI adoption while prioritizing security. Companies across the cybersecurity landscape are mobilizing efforts to create secure and responsible AI environments, emphasizing the need for comprehensive strategies that can effectively mitigate the associated risks. The conference highlighted the unprecedented pace at which technology is evolving and the corresponding imperative for robust security measures to safeguard organizations as they integrate AI into their operational frameworks.