With this acquisition, Checkpoint will set new standards for cybersecurity and will be able to offer a complete end-to-end AI security stack designed to protect businesses as they accelerate their AI journey
Check out Point Software Technologies Ltd. Today, a pioneer and global leader in cybersecurity solutions, announced that it has concluded an agreement to acquire Lakera one of the world's leading AI-Native security platforms for agent AI applications. The acquisition will allow Checkpoint to set new standards for cybersecurity and provide a complete end-to-end AI security stack designed to protect businesses as they accelerate their AI journey.
“AI is transforming all business processes, but we are also introducing new attack surfaces. We chose Lakera for its native AI security, excellent accuracy and speed at scale.
– Nadav Zafrir, CEO, Checkpoint Software Technology
Companies are rapidly incorporating large-scale language models, generator AI, and autonomous agents into their core workflows. This accelerates innovation, but expands the attack surface, from data exposure and model manipulation to risks introduced through multi-agent collaboration and autonomous decision-making. Once data became new executables and the Internet of agents were formed, real-time AI-Native security became a business mandate.
Checkpoints already secure this transformation through Genai Protect, SaaS and API security, advanced data loss prevention, and machine learning-driven defenses for applications, clouds, and endpoints. With Lakera, Check Point extends these capabilities to provide one of the industry's first end-to-end AI security stacks. Combining Lakera's runtime protection with AI-powered Checkpoint Infinity architecture, businesses can ensure the full lifecycle of AI (models, agents, data).
“Lakera was dedicated to an AI era centered around real-time runtime security and research. By participating in checkpoints, you can accelerate and expand your mission globally. Protect LLMS, generator AI, and agents with speed, accuracy and guardrails.
-David Haber, co-founder and CEO of Lakera
Founded by AI experts from Google and Meta, Lakera was designed specifically for the AI-Native environment. The company operates major AI R&D centers in Zurich and San Francisco. Its flagship solutions, Lakera Red and Lakera Guard, provide pre-deployment attitude assessment and real-time runtime enforcement to protect LLMS, AI agents, and multimodal workflows. Lakera's platform combines advanced runtime protection with a series of red teams enhanced by Gandalf, the global leader in hostile AI networks. This unique approach guarantees evolving defenses that go beyond the new AI threats, giving businesses the confidence to deploy AI at a safe and large scale. Lakera is trusted by Fortune 500 Enterprises around the world and features a team of 11 PhDs, ensuring both technical rigor and corporate-grade resilience.
Key strengths of the Lakera platform:
- ai-native protection: Built from the start for AI, Lakera protects agents with LLMS, generation AI, and prompts, RAG, and MCP, providing real-time defense against rapid injection, data leaks, and model operations.
- Proven performance on scale: It offers 50ms latency and detection rates above 98%, with false positives below 0.5%, allowing businesses to protect AI-driven workloads without affecting speed or accuracy.
- Continuous Intelligence: With hostile patterns of over 80 million Gandalf, led by a dedicated AI research team, Lakera's platform is constantly adapting to the new AI threats.
- Global coverage: Supports over 100 languages and enables businesses to protect worldwide
Once closed, Lakera will form the foundation for the global center of excellence of checkpoints for AI security, accelerating the integration of the entire AI security investigation, innovation, and checkpoint infinity platform secure AI-powered companies and establish checkpoints as a global leader in shaping the future of AI security.
The transaction is expected to close in the fourth quarter of 2025 subject to customary closure conditions.
