Check Point and NVIDIA partner to secure enterprise AI

Check Point and NVIDIA are working together to protect AI infrastructure and help enterprises deploy AI safely. Check Point’s AI Cloud Protect is now part of NVIDIA’s Enterprise AI Factory Validated Design. It provides runtime cybersecurity, enhanced protection, and infrastructure acceleration. The system has been validated on NVIDIA RTX PRO servers. It can protect AI factories, which are data centers built for AI, without slowing down AI performance.

AI adoption is growing fast. Attacks are growing too. Enterprises and service providers are harnessing AI technologies and systems. Cybersecurity has become necessary throughout the AI pipeline covering every aspect from concept to usage of the system. Even early adopters are seeing attacks. AI infrastructure is being targeted. Prompts are being manipulated. Gartner reported that 32 percent of organizations experienced attacks involving manipulated prompts in the past year. 29 percent experienced attacks on generative AI infrastructure. Nearly 70 percent of cybersecurity leaders say generative AI risks are forcing them to rethink their traditional cybersecurity approach. Lakera found only 19 percent of organizations are very confident in their AI security. 49 percent are very concerned about vulnerabilities.

Check Point aims to cover the full AI supply chain. Infrastructure, networks, enterprise applications, and end users are all protected. AI Cloud Protect integrates with NVIDIA’s BlueField platform. It allows real-time monitoring across the AI infrastructure. It strengthens isolation between AI workloads. It gives deep visibility and control over AI data. This helps fight cyber threats and vulnerabilities without hurting performance as data volumes grow.

こちらもお読みください： クロスポイント・ソリューションズ、セキュリティ運用自動化サービスを発表

AI Cloud Protect uses エヌビディア DOCA Argus telemetry and Check Point’s AI-driven cybersecurity. Networks and hosts are protected in real time. The infrastructure layer runs on BlueField and does not use GPU resources. That means the AI factory is protected without slowing AI workloads.

At the application layer, CloudGuard Web Application Firewall protects AI applications. It stops prompt injection, jailbreaking, and LLM poisoning. Runtime protection covers LLM inputs, outputs, and all data flows, including on RAG and MCP servers. チェックポイント uses Gandalf, the world’s largest AI red teaming platform with over 80 million attack patterns, to detect attacks with very low false positives. This protects agent-based AI applications that interact autonomously with enterprise systems.

For users, GenAI Protect manages employee AI use. By doing so, it stops the leaking of confidential information and keeps the company’s activities within the bounds of the law. Employers are able to monitor the AI applications that their staff are using. Sensitive data is blocked in real time. Audit trails are created automatically for regulatory compliance.

On the network side, Quantum and CloudGuard Network Security firewalls protect generated AI. MCP servers and traffic are auto-discovered. AI tools and apps in use are identified. Traffic logs are created for analysis. Companies can manage all applications, including AI applications. Policies can be set to allow or block access. Agent-based AI, MCP servers, and shadow AI applications are fully manageable.