Zurich-based Lakera AI, on Thursday, October 12, announced the formal launch of its AI security platform, enabling developers to secure generative AI applications in the enterprise.
In addition to the launch, Lakera AI has secured $10M (approximately €9.4M) in a fresh funding round led by Redalpine.
Other founders, including Fly Ventures, Inovia Capital, and several industry expert angels, such as the co-founders of Snyk and key executives from Palo Alto Networks, Datadog, and Hybris , also participated.
Lakera AI: Secure generative AI applications
After a decade of developing machine learning systems for aerospace and healthcare, David Haber, Matthias Kraft, and Mateo Rojas-Carulla saw the rapid rise of LLMs. The trio knew existing enterprise security solutions would not accommodate the unique needs of AI.
They launched Lakera AI to fill this gap.
“When you build AI for airplanes and healthcare, high security standards become baked into everything you do. For AI to reach its full potential, we need these same high standards applied in the enterprise,” says David Haber, Founder and CEO of Lakera.
“Developers have a huge role to play in the AI era. AI security has to be incorporated into the development lifecycle. Enterprise developers need an intelligent compute layer embedded in AI applications to ensure protection over time. That’s what Lakera delivers,” adds Haber.
Lakera AI works with any AI model or modality and instantly provides enterprise-grade security for AI applications.
In August 2023, Lakera AI released the beta version of Lakera Guard, to meet the urgent demand for a solution that protected LLMs in production.
Lakera Guard is powered by a proprietary database that combines crowdsourced insights from Gandalf, the company’s viral AI education game, with open source databases and its own dedicated research.
Lakera’s database has amassed nearly 30M attack data points and is growing by 100k+ entries every day.
Lakera Guard empowers developers to activate this vulnerability data with one line of code, instantly embedding a security layer into AI applications in production.
Lakera Guard is LLM-agnostic and currently focuses on securing Q&A use cases by protecting against prompt injections (including jailbreaks, escape characters, role play and system prompt leakage), PII detection and masking, content moderation, and hallucinations.
“As enterprises are quickly establishing their generative AI stacks, we aim to be the go-to-choice for all things security across the organisation,” adds David Haber. “From developers to CISOs, it’s important that all stakeholders are empowered to integrate AI without the risks.”
Redalpine is a Swiss venture capital firm that brings together financial investment, operational expertise, and a vast international network to help ambitious entrepreneurs.
With over €1B in assets under management and a sector-agnostic investment strategy, the firm has backed some of Europe’s most disruptive companies, including N26, Taxfix, Inkitt, 9fin, Carvolution, Zenjob, vivenu, and Umiami.
Redalpine has over 85 companies in its portfolio and invests Europe-wide from its offices in Zurich and Berlin.
“AI presents security risks unlike anything we’ve seen before. Every organisation looking to leverage AI will have to navigate this new threat landscape,” says Harald Nieder, General Partner at Redalpine, Lakera’s lead investor. “This is one of the most experienced teams in AI with a treasure trove of vulnerability data. Their developer-first approach provides the missing link for all enterprises to deploy AI models with confidence.”