Protect AI builds an AI Security Posture Management (AI‑SPM) platform that secures ML models, LLMs, and AI applications with products such as Guardian, Layer, Radar, and Sightline. The platform has expanded from one to five products and includes SydeBox, an automated GenAI red‑teaming tool acquired via SydeLabs. Protect AI serves private and public sector customers, including national security organizations and Fortune 500 companies, and its huntr community exceeds 15,000 members. The company reports five open‑source offerings downloaded millions of times, MLSecOps.com leadership in AI security education, and 300% year‑over‑year team growth with plans to add 50 employees by the end of 2024. Protect AI will use the new funding to expand customer success and sales, accelerate R&D, and strengthen channel programs to extend its market lead. Financially, Protect AI closed a $60M Series B and has raised $108.5M to date after a $35M Series A one year earlier. Protect AI offers a suite of tools to secure AI systems, with a flagship product, AI Radar, that generates a machine learning bill of materials (MLBOM) by inspecting training data, test datasets, code and pipelines. The company also provides protections against prompt-injection attacks, scans Jupyter notebooks for issues like exposed credentials and PII, and flags problematic open-source licenses. Protect plans to use its new funding to add capabilities to its platform, expand its research effort and launch new open source projects, and intends to grow headcount from 25 to 40 by year-end. The startup says it has private- and public-sector customers across financial services, healthcare, life sciences and energy. Protect positions itself against peers such as Resistant AI, HiddenLayer, Robust Intelligence, CalypsoAI and Troj.ai. Financially, the company has raised a Series A and earlier seed capital, bringing total funding to $48.5 million. Protect AI develops security tooling to defend AI systems and machine learning models from exploits, with a flagship product called NB Defense that scans Jupyter Notebooks for credentials, PII, nonpermissive open‑source licenses and offers remediation suggestions. The company plans to extend support beyond notebooks to common AI development platforms including Amazon SageMaker, Azure ML and Google Vertex AI Workbench. Protect AI emerged from stealth roughly a year after being co‑founded by Ian Swanson, Daryan Dehghanpisheh and Badar Ahmed; Swanson and Dehghanpisheh previously worked together at AWS and Swanson worked with Ahmed at Oracle. NB Defense is available for free initially, with paid options to be introduced in the future. Protect AI is pre‑revenue, has about 15 employees, and says it has secured enterprise customers including firms in the Fortune 500 across multiple verticals. Management says the new capital will fund product development, customer outreach and hiring in software development, engineering, security and go‑to‑market roles.

Total raised

$109M

Funding rounds

3

Latest round

Series B

Latest activity

Aug 2024