Lakera builds security infrastructure that sits between an application and its language model to catch prompt injection, jailbreak attempts, and sensitive-data leaks before they reach a response. Founded in 2021 in Zurich by David Haber, Mateo Rojas-Carulla, and Matthias Kraft, the company built its reputation partly through Gandalf, a public game that challenges players to extract a secret password from an AI, which has since generated tens of millions of adversarial prompts feeding Lakera's threat research.
The core product, Lakera Guard, screens prompts and model outputs through a low-latency API call, and Lakera Red performs automated red-teaming against a customer's own AI systems to surface vulnerabilities before attackers do. In September 2025, Check Point Software acquired Lakera in a deal reported around $300 million, folding Guard, Red, and Gandalf into Check Point's broader AI security stack while keeping the Lakera brand and platform live.
Lakera suits engineering and security teams shipping LLM-powered products who need a dedicated guardrail layer rather than building detection in-house. A free Community plan covers up to 10,000 requests a month; beyond that, Pro and Enterprise pricing is set per contract through Check Point's sales process.
