Universal Cognitive Architecture (UCA) is a standard for building AI systems that use an organisation’s stored knowledge, rather than relying on the built-in knowledge of a large language model (LLM).

UCA is an open standard that is permanently free under Creative Commons, allowing organisations to build on it without a licence or permission. The UCA allows an organisation to retain ownership over its own knowledge in plain text.

The model works alongside the model context protocol (MCP), which is an open standard created by Claude owner Anthropic for connecting AI applications and LLMs to external data, tools and systems.

While MCP helps connect an AI model to a repository, UCA aims to help the model find the right source quickly. 

Together, the two support a Domain Language Model, which is a stateless model reasoning over a sovereign repository. This means an organisation can use the best AI model for each job while retaining data sovereignty.

Chris Kincade, chief executive and founder at Starling Memory Works, said: “The current conversation on AI sovereignty focuses on infrastructure – where models run, where data sits, and who owns the compute. It misses the real story.

“Sovereignty is no longer a trade-off. Governed memory produces better intelligence than what AI generates. It detects drift and keeps every conversation current – giving businesses the very real gift of context.”