Lakera Security Platform
Lakera Security Platform protects enterprise AI deployments by detecting and preventing prompt attacks, data leakage, and unauthorized AI use in real-time. It secures employee AI usage through shadow discovery and policy controls, protects AI agents with runtime threat detection, and identifies vulnerabilities through risk-based red teaming. Trusted by Fortune 500 companies like Dropbox and AWS, Lakera operates with sub-50ms latency across any LLM model and reduces security risks by 3–4 orders of magnitude.
Lakera Security Platform is an AI-native security solution designed for enterprises deploying large language models and AI agents at scale. It addresses three core security challenges: detecting and preventing prompt attacks and data leakage in real-time, discovering and controlling shadow AI usage across employee workflows, and identifying vulnerabilities through risk-based red teaming. The platform operates model-agnostic and multimodal, securing chatbots and audio bots across any LLM with sub-50ms latency even for large prompts and context windows. Key capabilities include shadow AI discovery across apps and browsers, context-aware data protection in prompts, granular policy controls by user and action, runtime threat detection for AI agents, and collaborative vulnerability remediation guidance. Lakera's context-aware approach reduces security risks by 3–4 orders of magnitude and scales from zero to hundreds of prompts per second without requiring code changes. Trusted by Fortune 500 companies including Dropbox, AWS, Asana, and Nu Bank, the platform is recognized as a representative GenAI TRiSM vendor in Gartner's 2024 Innovation Guide and addresses OWASP's Top 10 LLM risks. It offers flexible deployment options—SaaS or self-hosted—with enterprise support, SSO, RBAC, SIEM integration, and data residency choices for regulated environments.
The terms do not explicitly mention AI training on user content. However, Section D grants Check Point broad rights to u
No training opt-out mechanism was identified in the legal documents.
High risk across multiple dimensions. Only use with public, non-sensitive data.
The terms do not explicitly mention AI training on user content. However, Section D grants Check Point broad rights to use submitted Feedback for 'any purpose' including creating derivative works, which could potentially include training purposes. The language is vague enough to create moderate risk.
“you grant Check Point a perpetual, irrevocable, worldwide, royalty-free, transferable, and sublicensable right to use, modify, reproduce, distribute, display, and create derivative works from such content for any purpose”
Section D grants extremely broad license rights to Check Point over user submissions. The license is perpetual, irrevocable, worldwide, royalty-free, transferable, and sublicensable for 'any purpose without limitation.' This is one of the broadest possible content licenses.
“perpetual, irrevocable, worldwide, royalty-free, transferable, and sublicensable right to use, modify, reproduce, distribute, display, and create derivative works from such content for any purpose without notice or compensation”
Section D explicitly states Check Point can authorize third parties to use submitted content ('to authorize others to do so as well'). However, general third-party sharing practices are not detailed. The Privacy Policy is referenced but not provided in this document.
“to authorize others to do so as well. You are prohibited from posting or transmitting to or from this Site any unlawful, threatening, libelous, defamatory, obscene, pornographic”
The terms do not specify data retention periods or deletion rights. Section D indicates Check Point can use Feedback perpetually with no obligation to disclose use. The Privacy Policy is referenced but not included, creating uncertainty about actual retention practices.
“Check Point shall have no obligations to you with regard to its use or disclosure of such Feedback, and Check Point shall have a worldwide, perpetual, irrevocable and fully-paid right to use”
Section D is extremely unfavorable to users. Check Point claims ownership/control of all user-submitted content and Feedback with perpetual, irrevocable rights. Users retain no meaningful IP ownership over their submissions. This is a high-risk provision.
“By submitting Feedback or other content, you grant Check Point a perpetual, irrevocable, worldwide, royalty-free, transferable, and sublicensable right to use, modify, reproduce, distribute, display, and create derivative works”
The terms do not mention SOC2, GDPR, HIPAA, ISO certifications, or other compliance certifications. Section B references a Privacy Policy but provides no information about compliance standards or certifications. This creates moderate uncertainty about enterprise compliance posture.
“Any personal data shared through the website will be processed in accordance with Check Point's Privacy Policy (https://www.checkpoint.com/privacy/)”
Track and measure your AI tools with Olakai's observability platform.
Explore Olakai