Netskope launches unified platform for AI security

Netskope has launched a suite of AI security tools it says gives organisations a single place to manage risk across AI agents, AI applications, user activity and the data that flows between them.

The new offering, called Netskope One AI Security, sits within the broader Netskope One platform. It targets a range of AI-related risks that security teams have flagged as the use of generative AI and agentic workflows spreads across businesses.

Netskope described the update as a response to gaps created when staff and contractors use AI tools outside formal oversight, often referred to as Shadow AI. It also pointed to concerns such as data loss, prompt injection and jailbreaking, which can manipulate a model's behaviour or extract sensitive information.

The suite includes four products: Netskope One Agentic Broker, Netskope One AI Guardrails, Netskope One AI Gateway and Netskope One AI Red Teaming. All four are generally available.

Agentic controls

Netskope One Agentic Broker focuses on visibility and control for agentic AI workflows, including approaches to securing agentic AI before it scales. It is designed to track and secure Model Context Protocol transactions, including sanctioned and unsanctioned activity.

Model Context Protocol, often shortened to MCP, is emerging as a way for AI agents to access tools and data sources. As organisations connect agents to internal systems, MCP traffic can become another pathway for data access and potential leakage.

The broker decodes and secures MCP traffic between AI agents and enterprise data sources, and provides monitoring and governance for how agents interact with those sources.

Prompt risks

Netskope One AI Guardrails targets threats that arise from user prompts and automated interactions. It is designed to prevent prompt injection and jailbreaking attempts, and to provide content moderation for large language model interactions.

According to Netskope, the product integrates with Netskope One Data Loss Prevention and Threat Protection, giving security teams a single view of detections and related context rather than separate alert streams.

It also blocks content Netskope categorises as discriminatory, copyrighted or inappropriate, which it says reduces legal and reputational risk. Netskope added that detections are mapped to MITRE ATLAS and the OWASP Top 10 for LLMs.

Private AI inspection

Netskope One AI Gateway targets private AI environments, including on-premises and virtual private cloud deployments, with a focus on protecting private AI from development to runtime. These settings can fall outside the reach of traditional cloud inspection tools, limiting monitoring and enforcement.

According to Netskope, the AI Gateway extends controls to traffic that does not go through the cloud for inspection. It centralises authentication, traffic management and content inspection for agentic communications in private AI environments.

It also secures interactions between applications and internally hosted large language models, and applies governance to data flows generated by autonomous agents.

Testing models

Netskope One AI Red Teaming is aimed at earlier stages of AI deployment. It simulates attacks against models and AI applications and assesses safety risks.

Netskope said the product exposes AI models to thousands of simulated attacks and identifies complex multi-turn attacks that can bypass default guardrails. It also said the process can detect performance drift linked to vulnerabilities.

Market context

The launch comes as enterprise spending on AI continues to rise. Netskope cited IDC figures putting worldwide enterprise AI spending at USD $241.8 billion in 2025, with projections of more than USD $867.3 billion by 2029.

Netskope argued that rapid adoption has created security blind spots, particularly for organisations relying on older security tools. It said those products can struggle to track how data is used across AI applications, private and public models, and AI agents and tools.

It also highlighted risks that can arise when AI systems bypass existing controls, including data exfiltration, manipulative prompts and inappropriate usage.

Network layer

Alongside the AI security tools, Netskope highlighted a set of network features it calls NewEdge AI Fast Path, which it said optimises latency when customers connect to AI services.

Netskope framed the combination of network routing and security policy as a way to balance performance and protection as AI usage expands. It said the approach uses context-aware controls tied to activity observed across its platform.

"The AI Supercycle is here, demanding a new standard for high-performance security and networking. We believe the next decade will be defined by an intelligent edge. Our Netskope One platform combined with our NewEdge network is exactly that: a structural architecture built specifically for the requirements of an autonomous, agentic economy. This architecture provides an AI-native foundation with the innate ability to secure the complex data flows that power the modern AI ecosystem," said Sanjay Beri, Co-Founder and CEO of Netskope.

Colgate-Palmolive is among the customers working with Netskope on AI-related data protection, according to a quote included with the announcement.

"Colgate-Palmolive's 220-year history of innovation is the foundation of our company's purpose as a caring, innovative growth company that is reimagining a healthier future for all. We leverage technology solutions, including Artificial Intelligence, to drive creativity, efficiency and scale. AI systems offer powerful means for enhancing our business processes, improving customer and consumer experiences, and driving innovation across our operations. That said, technology comes with responsibilities, which is why we are guided by our employee Code of Conduct and AI Governance structure. We ensure the responsible deployment and use of AI systems by prioritising good judgment, adherence with applicable laws and regulations, and ethical considerations to mitigate any unintended consequences. Colgate-Palmolive continues to partner with Netskope to provide data protection associated with our AI systems and associated use cases," said Alexander Schuchman, CISO and SVP Global Network Engineering and Operations, Colgate-Palmolive.