Source of truth for AI
A source of truth for AI is the authoritative knowledge base an AI agent reads from, kept current and permission-aware so every agent gets the same accurate answer.
Category
Industry concept
See all terms
What is a source of truth for AI?
A source of truth for AI is the place an AI agent treats as authoritative when it needs a fact about the company. It's the same idea as a single source of truth, one place holds the answer instead of several disagreeing copies, but built for a reader that isn't a person. The source the agent reads from has to already be resolved, current, and scoped to what that agent is allowed to see.
How does a source of truth for AI differ from a single source of truth?
A single source of truth is a business principle: pick one authoritative place for a fact so people asking the same question get the same answer. A source of truth for AI adds requirements a human reader doesn't need. It has to be machine-readable and structured enough for an agent to retrieve directly, it has to update continuously since agents run constantly, and it has to apply permissions per agent and per request.
How does a source of truth for AI work?
A source of truth for AI resolves the same fact appearing in multiple connected data sources into one current record, rather than leaving an agent to pick between conflicting versions. It keeps that record up to date through continuous updates as the underlying sources change, and applies access controls at the point of retrieval, so what one agent or Client can read matches its permissions rather than the widest access available in the repository. Qontext's context repository is built to serve this role directly: one graph-based, versioned store that every connected agent and Client reads from.