Since macOS 13 (2022), every Mac ships with a pair of daemons whose function is to construct a detailed personal knowledge graph. This graph maps user identity, relationships, communication patterns, locations, and activities. The existence and full functionality of these daemons are not officially documented by Apple.
The personal knowledge graph is built on-device by cross-referencing data from multiple sources, including Contacts, Calendar, Photos, Messages, Mail, location services, and Screen Time. It also incorporates a range of behavioral signals. The system infers relationships and activities based on this aggregated data.
The constructed graph is maintained as a queryable store, which is reportedly accessed by Large Language Models (LLMs). While Apple researchers published a paper in 2023 on the techniques behind graph construction, there is limited information on its integration into macOS, specific data collected, or the inference of 'kinship' relationships.
The daemons were encountered during security research on the Apple Intelligence bundle in macOS 26. Separately, Nic Lydon also identified the same database and documented it in a personal knowledge graph context. The only official mention is a brief man page entry for 'knowledgeconstructiond'.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's briefNew every morning, and the back catalogue is archived by date.
Every Mac running macOS 13 or later contains undocumented daemons that build a personal knowledge graph about the user and their contacts. This graph cross-references data from various applications and behavioral signals to infer relationships and activities. The system is reportedly queryable by LLMs and lacks official documentation regarding its full scope and integration.