AI & Data Privacy

Last updated: September 19, 2026

LedgerBeaver LLC, a Texas limited liability company, provides a hosted finance workspace where you bring your own AI. You choose which AI can access your books, or whether any third-party AI accesses them at all. This page explains, in plain English, who sees what.

1. Who sees your data

2. Choose your privacy level

What your AI provider does with the data it sees depends on how you connect. Three broad options, from most convenient to most private:

⚠️ Consumer subscription (Claude Pro, ChatGPT Plus, Gemini)

Depending on your plan and settings, consumer AI products may use your conversations to train their models. Most providers let you opt out — check your provider's data settings before connecting business books. Best for: personal use and testing.

✅ API key (Anthropic API, OpenAI API)

API and business plans often provide stronger data-use controls than consumer subscriptions, but terms, defaults, retention, and opt-out settings vary by provider and can change. Verify your provider's current business-data terms and account settings before sending financial data. Using a separately governed business/API account is the recommended default for business use.

🔒 Local model (Ollama, LM Studio, vLLM, llama.cpp)

The model runs on your own hardware, so no third-party AI provider ever sees your books. Data flows only between your machine and your LedgerBeaver workspace. Best for: maximum privacy. Setup guide below.

3. Running LedgerBeaver with a local model

Local runtimes like Ollama and LM Studio run the model — they don't speak MCP themselves. So you connect through an MCP-capable agent that uses your local model and talks to LedgerBeaver:

your local modelMCP-capable agent → LedgerBeaver
Ollama / LM Studio / vLLM / llama.cpp → Continue · Cline · OpenCode → our MCP endpoint
  1. Install a local runtime. Ollama (ollama pull a model that fits your RAM — a capable tool-use model is strongly recommended for accounting work) or LM Studio (download a model, start the local server).
  2. Pick an MCP-capable agent that supports custom / local model endpoints — for example Continue, Cline, or OpenCode. Point its model setting at your local runtime (Ollama and LM Studio both expose an OpenAI-compatible endpoint on localhost).
  3. Connect the agent to LedgerBeaver. Generate an API key in the app (Connect AI page) and paste the ready-made MCP config snippet for your agent. Every snippet on that page auto-fills your key.

Honest scope: "local" here means the AI is local. Your books are still stored in your LedgerBeaver cloud workspace (that's what makes them available to your accountant, your team, and your other devices). What the local option removes is any third-party AI provider from the data path.

4. Practical ways to minimize exposure

5. Your provider's policies

Because your AI provider is your choice, their privacy policy applies to what they see. The current policies for the most common providers:

6. Our commitments

Whatever AI you choose, on our side the rules don't change: your data is yours, encrypted in transit and at rest, never sold, never used for advertising, and never used to train models. The full detail lives in our Privacy Policy, Data & Compliance statement, and DPA. Questions: support@ledgerbeaver.com.