SentinelOne, a pioneer in AI-driven cybersecurity, announced a sweeping new vision and product roadmap for securing the AI-powered enterprise at its annual conference, OneCon 2025. The company introduced its expanded AI Security portfolio designed to help organizations secure artificial intelligence tools, agents, and data systems, while advancing automation and autonomous security operations.
The announcements underscore SentinelOne’s commitment to enabling businesses to innovate with AI confidently—by securing AI assets and leveraging AI for defense. The strategy centers on providing the technology, intelligence, and services needed to accelerate safe AI adoption and minimize risk in an increasingly automated world.
“Enterprises should be able to embrace AI innovation without fear,” said Tomer Weingarten, CEO of SentinelOne. “To do that, organizations must master both Security for AI and AI for Security. At OneCon 2025, we’re not only envisioning the future of AI security—we’re delivering the tools and frameworks needed to make it a reality today.”
Cyber Technology Insights : Dropzone AI Joins Fortune Cyber 60 List of Fastest-Growing, Most Innovative Cybersecurity Firms
Introducing SentinelOne’s New AI Security Portfolio
AI models, prompts, agents, and data pipelines are emerging as new targets for attackers, while security teams face the growing challenge of managing data complexity and the velocity of AI risks. To address this, SentinelOne launched a dedicated portfolio focused on securing AI assets across all use cases.
The new portfolio features three generally available products and one beta release, covering AI use by employees, developers, and enterprises:
Prompt Security for Employees – Provides complete visibility and control over corporate use of generative AI tools across 15,000+ AI websites. It detects and mitigates Shadow AI usage and protects against inadvertent data exposure.
Prompt Security for AI Code Assistants – Safeguards generative coding environments by automatically redacting secrets, personal data, and intellectual property from code. A built-in Vulnerable Code Scanner prevents insecure or malicious AI-generated code from reaching production.
Prompt Security for AI Applications – Shields custom-built AI applications—from chatbots to automation systems—against threats like denial-of-wallet and remote code execution attacks.
Prompt Security for Agentic AI (Beta) – The first comprehensive solution for monitoring and governing autonomous AI agents built on the Model Context Protocol (MCP). It offers real-time risk scoring and governance for machine-speed decision-making.
AI-Ready Data Pipeline: Integrating Observo AI
Following its recent acquisition of Observo AI, SentinelOne has introduced an integrated AI-native data pipeline within the Singularity Platform. This enhancement redefines how enterprises collect, enrich, and operationalize security data for agentic and autonomous operations.
The Observo AI Integration with Singularity™ AI SIEM brings together intelligent data streaming with advanced analytics and orchestration. The system ingests and normalizes petabytes of data in real time, applying pre-ingestion analytics to prioritize and streamline inputs for the Singularity AI SIEM. The result is the only SIEM capable of delivering both flexible pull/stream data collection and real-time, AI-powered insights for faster threat detection and response.
Cyber Technology Insights : SentinelOne and Google Cloud Launch Wayfinder to Redefine Managed Threat Detection in the AI Era
Next-Generation Purple AI: Expanding Agentic Capabilities
At OneCon 2025, SentinelOne also showcased major updates to Purple AI, its industry-leading agentic AI security analyst. These new capabilities combine human-level reasoning with autonomous investigation, helping analysts move from detection to resolution in minutes instead of hours.
Key innovations include:
In-line Agentic Auto-Investigations (Preview) – Fully automated, reasoning-driven investigations that analyze alerts, validate hypotheses, assess impact, and propose mitigation strategies. Purple AI documents every step of its analysis for human review, shifting the paradigm from “human-led AI assistance” to “AI-led human validation.”
Integration with Singularity Hyperautomation (Preview) – Enables seamless automation of pre-approved workflows, such as initiating communications with analysts via Slack or executing validated responses autonomously.
Agentic Custom Detection Rule Creation (Preview) – Allows analysts to instantly generate and deploy new detection rules recommended by Purple AI, enabling proactive defense against similar attacks.
These advancements empower security teams to move beyond traditional SOC models, combining reasoning, orchestration, and continuous learning to combat sophisticated threats at scale.
Cyber Technology Insights : Cyera Named Among Top 10 Fastest-Growing Companies, Leading the Future of AI and Data Security
Extending Purple AI Beyond the SentinelOne Ecosystem
SentinelOne is also opening its agentic security capabilities to third-party and custom AI applications. The company introduced the Purple AI Model Context Protocol (MCP) Server, designed to integrate securely with any large language model or AI framework. Acting as a universal translator, it connects external AI systems with the Singularity Platform, enabling developers and partners to build custom agentic workflows enriched by SentinelOne’s analytics and threat intelligence.
The open-source Purple AI MCP Server is available now on GitHub, marking a major step toward making AI-powered cybersecurity accessible and extensible across the broader ecosystem.
A Vision for AI Security Today
Through these innovations, SentinelOne is redefining what it means to secure AI-driven enterprises. The company’s AI Security framework combines proactive defense for AI systems with intelligent automation for security operations—empowering defenders to maintain visibility, control, and speed in a rapidly evolving digital landscape.
Cyber Technology Insights : Pentera Acquires AI Red Teaming Leader EVA Information Security
To participate in our interviews, please write to our CyberTech Media Room at info@intentamplify.com
