AWS enters the context layer race with a graph that learns from agents, not manual curation

AWS announced a three-part context intelligence stack for AI agents, led by AWS Context, a service that automatically builds and updates knowledge graphs from enterprise data. The company says the graph can infer relationships across datasets, business rules, and domain knowledge, then improve over time based on which sources agents use and trust.

AWS is pushing into the emerging context layer market, where vendors are trying to make enterprise data usable for agentic AI without costly custom integration. Its bet on self-learning graphs could reduce maintenance work for customers, but it also raises the bar for accuracy, governance, and transparency as agents rely more heavily on automatically inferred business context.