"More and more organizations are integrating large language models, generative AI, and autonomous agents into their business processes. While this accelerates innovation, it also creates new security challenges. In a world where data increasingly functions as "executable code," data breaches, model manipulation, and undesirable effects of autonomous decision-making are becoming ever greater threats. Check Point already offers GenAI Protect, SaaS and API security, data loss prevention, and machine learning-driven security. Adding Lakera's technology creates a more complete AI security stack."
"The technology combines advanced runtime security with continuous red teaming. This is supported by Gandalf, a global network for adversarial AI. This approach yields a dynamic security model that evolves in response to emerging AI threats. Lakera's platform demonstrates strong performance with a detection rate above 98 percent and latency below 50 milliseconds. The system generates less than 0.5 percent false positives, which, according to the manufacturer, ensures secure AI workloads without delay or inaccuracy."
Check Point acquired Lakera to build an end-to-end AI security stack that helps organizations deploy AI safely and responsibly as new attack vectors emerge. Organizations increasingly integrate LLMs, generative AI, and autonomous agents into business processes, creating new security challenges as data functions as "executable code." Lakera offers AI-native runtime protection with products Lakera Red and Lakera Guard, providing posture assessments and real-time protection for LLMs, AI agents, and multimodal workflows. The platform pairs advanced runtime security with continuous red teaming supported by Gandalf, a global adversarial AI network. Reported performance shows >98% detection, <50ms latency, and <0.5% false positives.
Read at Techzine Global
Unable to calculate read time
Collection
[
|
...
]