Zurich-based Lakera AI announced the official launch of its AI security platform, allowing developers to protect generative AI applications in the workplace. Along with the launch, Lakera AI has raised $10 million (about €9.4 million) in a new fundraising round managed by Redalpine.
Zurich-based Lakera AI announced the official launch of its AI security platform, allowing developers to protect generative AI applications in the workplace.
In addition to the launch, Lakera AI has raised $10 million (or €9.4 million) in a fresh fundraising round run by Redalpine.
Lakera AI was introduced to close this gap.
“High security requirements get built into everything you do when you develop AI for healthcare and airlines. According to David Haber, founder and CEO of Lakera, “We need these same high standards implemented in the workplace for AI to realize its full potential.
“In the age of AI, developers will play a critical role. Security for AI must be considered throughout the development process. Enterprise developers require an intelligent computing layer incorporated into AI applications to provide long-term safety. Lakera offers that, adds Haber.
Lakera AI quickly offers enterprise-grade security for AI applications and integrates with any AI model or methodology.
The beta version of Lakera Guard, developed by Lakera AI, was made available in August 2023 in response to the pressing need for a safeguard for LLMs already in use.
A unique database that blends open-source datasets, specialized research, and crowdsourced inputs from Gandalf, the company’s popular AI teaching game, powers Lakera Guard.
Nearly 30 million attack data points have accumulated in Lakera’s database, adding 100k+ entries daily.
With only one line of code, Lakera Guard enables developers to rapidly add a security layer to production AI systems by activating this vulnerability data.
Lakera Guard, which does not care about LLM, protects against PII detection and masking, content moderation, hallucinations, and prompt injections (like jailbreaks, escape characters, role play, and system prompt leaking). It also protects against PII and PII injections.
“We aim to be the go-to choice for all things security across the organization, as enterprises are quickly establishing their generative AI stacks,” says David Haber. “It’s critical that all stakeholders, from developers to CISOs, are empowered to integrate AI without the risks.”
Investor Redalpine is a Swiss venture capital company that supports ambitious entrepreneurs by combining financial investment, operational know-how, and a broad global network.
The company has funded some of Europe’s most innovative businesses, including N26, Taxfix, Inkitt, 9fin, Carvolution, Zenjob, vivenu, and Umiami, and manages over €1 billion in assets.
Redalpine invests across Europe from its headquarters in Berlin and Zurich and has approximately 85 firms in its portfolio.
“AI poses security dangers that have never been seen before. This new danger scenario will need to be navigated by every organization wanting to use AI, according to Harald Nieder, General Partner at Redalpine, Lakera’s principal investor. “With a wealth of vulnerability data, this is one of the most skilled AI teams. Their developer-first strategy fills the gap for all businesses to deploy AI models confidently.
