About Check Point
Check Point AI Security is a family of products for the AI systems an organisation builds and the AI tools its staff use. Check Point announced in September 2025 that it would acquire Lakera, and the Lakera products now carry the Check Point name. We offer three of them.
Check Point AI Red Teaming (formerly Lakera Red)
AI Red Teaming tests generative AI applications, agents and models by running adversarial attacks against them, then reports the results with a severity rating and remediation guidance. It runs as a web platform that Check Point operates, so there is nothing to install: you point it at a target, which can be a model, an agent endpoint or an API, review the scan plan and launch the run.
We resell the platform licence. Our own team also uses it in our AI penetration testing.
Check Point AI Guardrails (formerly Lakera Guard)
Your application calls AI Guardrails through an API to screen what goes into the LLM, the content it retrieves, its tool calls and tool responses, and what comes out. The API returns what it detected, and the application blocks or flags according to your policy. It detects prompt attacks, data leakage, malicious links and unsafe content, and has controls for agent behaviour.
You can use it as a managed SaaS service or host it yourself.
Check Point Workforce AI Security
Workforce AI Security watches how employees use AI tools in real time. It classifies the content and intent of each use and applies your policy: block, redact, restrict pasting or guide the user. It also discovers AI use that the organisation has not approved, often called shadow AI.
It deploys through a browser extension, and a desktop agent is also available.
Get Check Point through Incognito Lab
Incognito Lab is an official Check Point partner in Thailand.
Quotation
We prepare the quotation for your organisation directly.
Demo
Our team walks you through Check Point and answers your questions as you go.
Trial and proof of concept
Try it yourself, or run a proof of concept in your own environment before you commit.
Support
Once you are using it, our team handles your questions and support requests.
Frequently asked questions
Where can AI Guardrails run?
It runs as a managed SaaS service that processes requests in the EU, the USA or Singapore. Logs are stored in the EU by default, or in the US if you choose it when the organisation is created. It can also be self-hosted on Kubernetes or a Docker host, including in an air-gapped network. Self-hosting needs an AI Guardrails Enterprise licence.
Which AI tools does Workforce AI Security cover?
Check Point lists ChatGPT, Gemini, Claude, Copilot, GitHub Copilot and Cursor. Discovery also reaches web apps, desktop agents, browser extensions, SaaS integrations, code assistants and MCP servers, with an application catalogue that scores each tool's risk. The Enterprise tier covers web, desktop and IDEs, and the Essentials tier covers web applications.
What does AI Red Teaming test for?
Security, safety and responsible AI failures. Security objectives include instruction override, system prompt extraction, tool extraction and data exfiltration. The attacks are prompt injection, jailbreaks, multilingual and multi-turn conversations, and indirect attacks such as poisoned tool responses. Each result is rated Low, Medium, High or Critical by the share of harmful responses.
Which frameworks does AI Guardrails map to?
Check Point maps it to the OWASP Top 10 for LLM Applications (2025) and the OWASP Top 10 for Agentic Applications (2026). Its documentation also lists items with no runtime coverage, such as supply chain (LLM03).
Talk to us about Check Point
Ask us for a quotation, a demo, a trial or a proof of concept.
