Lakera offers a low-latency AI application firewall, Lakera Guard, designed to detect and block malicious prompts and other vulnerabilities in generative AI workflows. Its product uses a database of insights from open-source datasets, in-house ML research, and data collected via an interactive game called Gandalf to build a prompt injection taxonomy and continuously improve detectors. The company says its models work with any LLM, including OpenAI’s GPT-X, Google’s Bard, Meta’s Llama and Anthropic’s Claude, and it also provides specialized content-moderation detectors for hate speech, sexual content, violence and profanity. Integrations are provided via an API and a single-line code integration, alongside a centralized policy control dashboard for threshold tuning. Lakera launched in October after raising $10 million in initial funding and has since signed customers in North America including Respell and Cohere. With a fresh $20 million Series A, the company plans to expand its global presence, particularly in the U.S. Lakera sells an API-backed product called Lakera Guard that companies integrate into their applications to detect and mitigate prompt injections, data leakage, toxic outputs, and hallucinations. The company builds a database of attack patterns and signals from open-source datasets, in-house research, and an interactive game called Gandalf that has generated large-scale test data. Over the past six months Lakera recorded roughly 30 million interactions from about 1 million users, which it uses to develop a prompt-injection taxonomy and statistical defenses. Lakera says it already has major paying customers, including LLM developer Cohere, a leading enterprise cloud platform, and one of the world’s largest cloud storage services. The founders (including CEO David Haber) have advised on the EU AI Act, and the company says it will evolve the product as the threat and regulatory landscapes change. With $10 million in funding, Lakera is positioning to build out its platform following its public launch.

Total raised

$30M

Funding rounds

2

Latest round

Series A

Latest activity

Jul 2024