Privacy and data handling
AntiLarp acts as a processor for the reviewer who uploads documents. It restates claims at their verifiable information value; it does not build profiles of applicants or share documents with anyone other than the reviewer who submitted them.
What happens to a document
Uploaded resumes and applications are analyzed solely to produce the reviewer's report: claim extraction, tier assignment (verified, deflated, or unverifiable), and honest restatements. The analysis exists for the reviewer and no one else.
Storage and retention
Documents and their reports are stored so the reviewer can return to them, and are kept according to the instance's retention setting. Deleting a review purges its documents, reports, and queued work.
Model training
Documents and reports are never used to train models — neither by AntiLarp nor, per our provider terms, by the model API it calls.
External calls
Analysis makes exactly two kinds of outbound requests: calls to the model API that performs the restatement, and public-records lookups (such as scholarly indexes and code-hosting sites) whose queries are derived from individual claims. No document leaves the system in any other way.
Contact
Questions about data handling: nathan@nolarping.com.