MAIA Decision OS and Claude are two different things that work together. Understanding the distinction matters for data handling and trust.
MAIA Is the Engine
MAIA Decision OS is the decision rehearsal platform. It controls:
- The 9-stage pipeline that structures your rehearsal
- The 47 specialized agents that analyze different aspects of your decision
- The sequencing — which analysis runs when, what depends on what
- The compliance architecture — what data is screened, what language is permitted
- The artifact generation — your Decision Brief, Visual Intelligence report, and exports
- The data handling — Zero Data Retention, connector privacy, session lifecycle
Claude Is the Inference Engine
Claude (by Anthropic) processes the analytical reasoning within each agent. When MAIA's pipeline activates an agent — say, the one analyzing your dependency structure — it sends the relevant context to Claude's API and receives analytical output back. Claude does the thinking. MAIA controls what it thinks about, in what order, with what constraints.
Your Data and Claude
Your data flows through MAIA's pipeline and is sent to Claude's API for inference. Anthropic's commercial API policy states that data submitted through the API is not used to train their models. MAIA specifically uses API access (not consumer product access) to ensure this protection.
What MAIA Does NOT Do
MAIA does not provide recommendations or advice. MAIA rehearses your decisions, maps dependencies, simulates disruptions, and surfaces action paths. You decide which path to take. This distinction keeps MAIA in the decision support category — not the decision making category — which is where AI regulation is concentrating.