OpenLazarus: approval rules for AI agents that act on company records
An operating system with AI teammates that work within limits the company sets. Strong where a policy governs what AI does; not a monitor for staff use of outside chat tools.
Contents
| Type | AI operations platform |
|---|---|
| Headquarters | New York, US |
| Pricing | Published |
| Website | openlazarus.ai |
OpenLazarus is a company-owned operating system with AI teammates, each with an inbox, a phone number and limits the company sets1. For an AI policy, its relevance is the action clause: the company decides what runs automatically and what needs approval2.
Best for: Companies whose AI agents act on company records (customers, invoices, orders) and need approval rules and a visible record of every action.
Not the right fit for: Companies that only need to control what staff paste into ChatGPT, Copilot or Gemini, or that need a register of models built by a data-science team.
What OpenLazarus does for an AI policy
Its library lists 18 ready-made AI teammates, from bookkeeping and purchasing to customer support and IT helpdesk, plus the option to build a custom one; each works inside the company's own app on the same records staff use1. Each AI sees only its own database3, which narrows what any one agent can reach.
The approval rules are set per teammate. The manufacturing page gives the example of a material order over $20,000, or a change to a promised ship date, waiting for a person, and states that every action is logged where staff can see it4. The financial-services page states that money never moves until a person releases it, that client messages showing performance wait for approval and that every action is logged where compliance can see it5. In the terms of our template, that is clause 7 (always, ask, never) and clause 15 (records) enforced by the system.
The code and data belong to the company, which can host with OpenLazarus or move the system elsewhere, and European companies can run on servers in the EU2.
Strengths and limits
Strengths
- Approval rules sit inside the system the AI teammates work in, so the “ask first” clauses of a policy are enforced at the moment of action.
- Its industry pages give concrete limits: on the manufacturing page, a material order over $20,000 or a change to a promised ship date waits for a person; on the financial-services page, money never moves until a person releases it.
- Each AI teammate sees only its own database, which narrows what any one agent can reach.
- Code and data belong to the company, which can host with OpenLazarus or move the system elsewhere. EU-based companies can run on servers in the EU.
Limits
- It governs the AI teammates that run inside its own system, not staff use of outside chat tools such as ChatGPT or Gemini.
- No mapping to the EU AI Act, NIST AI RMF or ISO/IEC 42001 is described on the pages we read.
- No security certifications or named customer results appear on the pages we read.
- Prices are in US dollars, and engineering support starts at $4,500 a month.
Choose another option if: Your main risk is staff pasting confidential data into outside AI tools. Microsoft Purview or Google Workspace’s admin controls address that directly.
Pricing
Published: plans at $100 or $300 a month including the same amount in AI credits, managed hosting from $150 a month, engineering support from $4,500 a month and a 5.5% fee on AI credits.
Questions to ask OpenLazarus
- How are "ask first" rules configured for each teammate, and can a rule name a specific approver and a numeric limit?
- Where is the action log kept, how long is it retained and can it be exported for an audit?
- Which actions can be made impossible rather than only subject to approval?
- What happens to the log and the data if we move the system to our own cloud?
See how OpenLazarus compares with the other tools in the main guide.
Sources
- AI teammates. OpenLazarus. Accessed 8 Nov 2026.
- Frequently asked questions. OpenLazarus. Accessed 8 Nov 2026.
- OpenLazarus. OpenLazarus. Accessed 8 Nov 2026.
- OpenLazarus for manufacturing. OpenLazarus. Accessed 8 Nov 2026.
- OpenLazarus for financial services. OpenLazarus. Accessed 8 Nov 2026.
- Pricing. OpenLazarus. Accessed 8 Nov 2026.