Highlights:

  • Systems employing artificial intelligence and machine learning Protect AI has exited stealth mode with USD 13.5 million in new investment and the release of its first product, NB Defense.
  • It is estimated that there are currently over 10 million publicly available Jupyter Notebooks, with the number increasing by more than 2 million every year.

Protect AI, the Artificial Intelligence (AI) and Machine Learning (ML) systems cybersecurity startup, has exited stealth mode with raising USD 13.5 million in new investment and the release of its first product, NB Defense.

NB Defense is a free product that claims to be the first security solution in the market to address vulnerabilities in Jupyter Notebooks, a component utilized at the beginning of the ML supply chain. These web-based apps let developers generate and share documents with live code, equations, visualizations, and other data for coding purposes, such as data cleansing and transformation, statistical modeling, data visualization, and machine learning.

It is estimated that there are currently over 10 million publicly available Jupyter Notebooks, with the number increasing by more than 2 million every year. It is also suspected that there are several other Jupyter Notebooks installations in private repositories.

Protect AI was formed by a leadership team with AI business experience at Amazon Web Services Inc. and Oracle Corp., including co-founder and CEO Ian Swanson who was former AWS’s global head of AI and machine learning. Acrew Capital and boldstart ventures co-led the funding round, with Knollwood Capital, Pelion Ventures, and Avisio Ventures also participating.

Ian Swanson stated, “I have seen over 100,000 customers deploy AI/ML systems and realized they introduce a new and unique security threat surface that today’s cybersecurity solutions in the market do not address. This is why we founded Protect AI. ML developers and security teams need new tools, processes, and methods that secure their AI systems.”

Swanson noted that since virtually all ML code begins with a notebook, the business believed it to be the most natural starting point for accelerating a necessary industry shift.

Ian Swanson said, “We are launching a free product that helps usher in this new category of MLSecOps to build a safer AI-powered world, starting now. But many more innovations that will be released quickly across the entire ML supply chain.”

As MLOps has aided in accelerating the deployment of machine learning in production, the likelihood of security incidents has grown, and new vulnerabilities have been introduced into the enterprise machine learning supply chain. Examples of security vulnerabilities include Jupyter Notebooks incompatible with existing static code analyzers, tainted training data, arbitrary code execution in serialized models, and model evasion utilizing adversarial machine learning approaches.

NB Defense adds a layer of translation from existing security capabilities to enable scans of Jupyter Notebooks, then communicates results natively in the notebook or via reports with context-specific connections to problematic locations inside the notebook for rectification.

The offering examines a notebook for the standard Common Vulnerabilities and Exposures database in open-source ML frameworks, application tokens, libraries, and packages, and other credentials, and nonpermissive licensing in the frameworks.

NB Defense is currently accessible with a free license. Users can install NB Defense and utilize the JupyterLab Extension or Command Line Interface.