Factual drift occurs when a claim is cited, summarized, re-cited, re-summarized, and re-cited again across multiple media outlets — each AI-assisted summary introducing a small distortion until the final version of the claim is meaningfully different from the original. This drift is particularly dangerous in legal contexts, where a claim that originally referred to "estimated losses of $40M" may become "confirmed losses of $400M" through repeated AI paraphrasing.
Lawyers who cite media sources in legal documents need to verify not just that a claim was made, but that the version of the claim they're citing accurately reflects the original source. Omniscient AI helps by cross-checking the specific claim text against three AI knowledge bases — often revealing whether the claim is consistent with the widely-understood version of the fact or represents a drifted variant.
For litigation research, this drift detection capability has direct value. Cases where the opposing party has cited a drifted version of a fact in their pleadings can be challenged specifically and credibly — pointing to the drift and the original source simultaneously with support from the Omniscient AI verification record.