Check Point Software Technologies has launched the Maestro Hyperscale Orchestrator 200 (MHO 200), a network-security platform designed to inspect high-volume, largely encrypted AI traffic without compromising throughput.
The MHO 200 is aimed at neocloud providers, telecommunications operators, colocation companies, enterprises modernising data centres and governments building sovereign AI infrastructure.
It adds 400GbE ports and increases total fabric capacity to 12.8 Tbps, four times the 3.2 Tbps offered by the previous MHO 175, while retaining a 1RU form factor. Check Point also says threat-prevention throughput rises to as much as 4.5 Tbps per Maestro system, compared with 1.1 Tbps previously.
Security at AI scale
The launch responds to a central challenge in AI infrastructure: training, inference and agentic workloads can generate substantial east-west traffic within data centres, much of it encrypted. Organisations have often had to choose between inspecting that traffic and maintaining acceptable performance, or overprovisioning security hardware to accommodate demand spikes.
The Maestro product line is positioned as a hyperscale firewall orchestrator that combines firewalls of different models and generations into a single system that can expand with demand. The system can be divided into up to eight independent security groups, each with its own policy rules, without sacrificing advanced threat prevention.
“AI is forcing a rethink of network capacity planning,” said Pete Finalle, research manager for security and trust at IDC. “Organisations should not have to choose between inspecting that traffic and keeping it moving.”
He added that platforms able to scale threat prevention elastically, without overprovisioning hardware, would be essential for neoclouds, telcos and enterprises building AI infrastructure.
Asia-Pacific relevance
Although the announcement was made in Singapore, its relevance extends across Asia-Pacific, where investment in AI factories, sovereign AI platforms, GPU-as-a-service offerings and regional data-centre capacity is accelerating.
The MHO 200 is designed for environments where 400GbE is becoming a standard for high-end data centres and AI factories, including telecom operators protecting backbone networks and public-sector bodies pursuing data sovereignty.
For governments pursuing sovereign AI, Check Point says its AI Factory Firewall integrates with NVIDIA BlueField data-processing units. The platform can discover and secure AI applications, large language models, AI agents and Model Context Protocol servers directly from the firewall, while live patching applies security updates without reboots or downtime.
“AI is changing the scale at which networks need to be protected,” said Nataly Kremer, chief product officer at Check Point. “With the MHO 200, they can inspect all their traffic at AI speed, grow capacity as demand rises, and keep their operations running through failures.”
Resilience and operations
Check Point says the MHO 200 supports always-on resilience, with sessions remaining live through failures and automatic failover across two sites to deliver 99.999% uptime.
Its single-object management is intended to let security teams operate an entire multi-firewall system as one, while Agentic Network Security Orchestration automates policy tuning, health monitoring and compliance.
The MHO series builds on its prevention-first approach and follows earlier releases including the AI Network Firewall, Agentic Network Security Orchestration and AI Data Center Security.
