Fund

Lakera Raises $20 Million to Protect Enterprises From Potential Vulnerabilities in Large Language Models

Table of Contents

The Dark Side of Generative AI: Threats and Challenges

Generative AI has emerged as the poster child for the burgeoning AI movement, driven by popular apps such as ChatGPT. However, it remains a cause for concern within enterprise settings due to issues around security and data privacy.

What is Generative AI?

Large language models (LLMs) are the engines behind generative AI and enable machines to understand and generate text just like a human. These models can be trained on vast amounts of data, enabling them to learn patterns and relationships within that data.

The Threats: Malicious Prompts and Data Leakage

While LLMs have many benefits, they also introduce new risks. One such risk is malicious prompts, which are carefully crafted instructions designed to trick the application into doing something it’s not supposed to. This can include divulging confidential data that was used to train the model or giving unauthorized access to private systems.

Lakera: The AI-First Solution

Founded out of Zurich in 2021, Lakera is a Swiss startup that has raised $20 million in a Series A round led by European venture capital firm Atomico. Its express promise is to protect organizations from LLM security weaknesses such as data leakage or prompt injections.

How Lakera Works

At its core, Lakera is pitched as a "low-latency AI application firewall" that secures traffic into and out of generative AI applications. The company’s inaugural product, Lakera Guard, is built on a database that collates insights from myriad sources, including publicly available ‘open source’ datasets such as those hosted on Hugging Face, in-house machine learning research, and a curious interactive game called Gandalf.

Gandalf: The Interactive Game

Lakera’s Gandalf image credits LakeraThe game is designed to invite users to attempt to trick the application into revealing a secret password. As the levels progress, the game gets more sophisticated (and thus more difficult to "hack"). But these interactions have enabled Lakera to build what it calls a ‘prompt injection taxonomy’ that separates such attacks into categories.

Lakera’s AI-First Approach

"We are AI-first, building our own models to detect malicious attacks such as prompt injections in real time," said David Haber, co-founder and CEO of Lakera. "Our models continuously learn from large amounts of generative AI data, enabling them to adapt and improve over time."

The Benefits of Lakera’s Solution

Lakera’s solution provides several benefits, including:

Improved security: By detecting and blocking malicious prompts, Lakera ensures that applications are protected from potential threats.
Enhanced trust: With Lakera’s solution in place, organizations can have greater confidence in their ability to safeguard sensitive data.
Increased efficiency: By automating the detection and response process, Lakera frees up resources for more strategic tasks.

The Future of Generative AI: Challenges and Opportunities

As generative AI continues to evolve, it’s clear that new challenges will arise. However, with solutions like Lakera on the horizon, there is hope for a more secure and trustworthy future.

Conclusion

Lakera’s $20 million funding round is a significant step forward in the development of generative AI security solutions. With its AI-first approach and innovative technology, the company is well-positioned to address the growing threats associated with malicious prompts and data leakage.

By staying ahead of the curve and developing cutting-edge solutions like Lakera Guard, organizations can ensure that their applications remain secure and trustworthy.

Related Articles

Inside the wild fall and last-minute revival of Bench, the VC-backed accounting startup that imploded over the holidays
Generative AI funding reached new heights in 2024
Microsoft to spend $80 billion in FY’25 on data centers for AI