Palo Alto targets Portkey for AI agent security
The expansion of autonomous agents is reshaping security challenges in enterprise environments.
Palo Alto Networks announced today its intention to acquire Portkey to integrate AI control and protection capabilities.
Palo Alto Networks announced today its intention to acquire Portkey, a company focused on AI Gateways, in a context where enterprise adoption is advancing toward the use of autonomous agents.
The company indicated that Portkey provides a centralized control plane to manage and protect these systems, with the capacity to process billions of tokens per month with low latency.
The company noted that adoption is evolving toward autonomous agents, widening the AI security gap. This implies that these agents, operating as highly privileged insiders and executing automated decisions across internal and external systems, require broader control and protection mechanisms.
AI Gateway integration.
In this scenario, Portkey will serve as the AI Gateway for Prisma AIRS, acting as a system capable of monitoring, routing, and protecting each AI transaction across the enterprise.
Lee Klarich, Chief Product & Technology Officer at Palo Alto Networks, explains:
“As autonomous agents are integrated into the enterprise workforce, they also become a new unmanaged attack surface. By integrating Portkey into Prisma AIRS, organizations will be able to implement and govern AI agents with confidence. With Portkey, we are providing companies with visibility into all their agentic traffic and enabling them to control and protect against agentic threats.”
Unified security architecture.
In practice, the proposed architecture enables the consolidation of traffic inspection, security and governance policies, least-privilege access control, semantic routing, failover mechanisms, technical telemetry, and audit logs into a single operation.
The proposal includes securing AI interactions through inspection and real-time policies to identify threats and protect data, while implementing strict least-privilege controls in each interaction between agents to maintain security and compliance.
AI reliability and governance.
It also aims to ensure mission-critical reliability through availability of up to 99.99%, supported by semantic routing and automated failover, along with real-time visibility into each AI interaction.
The model also incorporates global AI governance through centralized artifact management, version control, and secure access across models, agents, and MCP servers.
This enables the transformation of fragmented environments into structured operations, while also reducing costs through caching techniques and granular quotas, with access to more than 3,000 LLMs and MCP tools from a unified interface.
Regarding this, Rohit Agarwal, CEO and co-founder of Portkey, states:
“Scaling AI in production requires a delicate balance between full flexibility for developers and absolute control for security teams. By joining Palo Alto Networks, we will establish the AI Gateway as the foundational layer of the secure AI enterprise. Together, we will provide the infrastructure that enables any organization to deploy autonomous agents with the confidence that their data and operations are fully protected.”
After the closing of the transaction, Palo Alto Networks will continue providing support, while the operation is expected to be completed in 2026, subject to customary closing conditions.
|
|
