Lazarus AI, a provider of production AI for heavily regulated organisations, has secured $43m to expand a platform that lets enterprises run advanced AI without giving up control of their data, models or costs.
The round drew backing from insurers and corporate investors alongside venture firms. Participants included QBE, Mucker Capital, Motivate Ventures, State Farm Ventures, Wave Function Ventures, Indelible Solutions, Joyance Partners, AllegisNL Capital, One Way Ventures, AngelList, Techstars and Plug and Play, among others.
Lazarus supports insurers, government bodies and technology businesses across three production use cases. Its Applied Intelligence System extracts findings from complex documents and links each one back to the page it came from. Across 150 customer use cases, the system has lifted field-level accuracy by 12.5 points compared with standalone models.
Its Clearwing product scans source code for vulnerabilities, reproduces them, assesses the supporting evidence and suggests remediation, while engineers retain the final decision on deployment. A third product, Constructor, extracts business logic from legacy systems and produces modern code shown to behave identically to the original. Two of the largest mainframe manufacturers are now piloting Constructor.
Customers can deploy the platform on their own premises, within sovereign environments or inside their own cloud. Lazarus charges no per-token fees, and customers do not need to rebuild anything when models or hardware are swapped out. The company said that in a modelled high-volume scenario, the platform reduced yearly operating expenses by 78% against frontier models billed on a pay-per-use basis.
Lazarus focuses on applying AI to high-stakes problems while leaving customers in charge of their data, models, deployment and spending.
The company plans to put the new capital towards broadening its strategic partnerships, increasing research and development, and launching additional products.
Lazarus AI CEO Alex Panait said, “The opportunity in AI is enormous. So is the question of control: who chooses the models, where data runs, how costs behave, and whether an organization can still change direction. Our customers keep sensitive data inside their walls, approve every model change and see the evidence behind every result.”
Copyright © 2026 RegTech Analyst
Copyright © 2026 RegTech Analyst
