How ArtMetrics Uses AI
ArtMetrics uses AI to assist with organization, summarization, and reviewer workflow support while keeping source evidence and human accountability central.
Narrative overlays
AI assists with summarizing public filings and long-form organization narratives into readable overlays, and with grouping signals, explanations, and source excerpts into consistent diligence workflows.
Source material
AI-generated narrative and summary text is always grounded in the underlying public filing or record it describes. Source filings and citations remain available alongside any AI-assisted summary so a reviewer can verify the original evidence.
Human correction
AI-drafted summaries and internal review notes remain subject to human review before publication or use. Reviewers can correct, override, or remove AI-assisted narrative text, and organizations can dispute it through the standard corrections process.
No autonomous scoring
LLMs are never used to calculate the 0-100 governance or financial scores displayed on ArtMetrics. The scoring engine is 100% deterministic, mathematically verifiable, and rule-based, built from IRS 990 compensation, revenue, and reserves data. AI is used only for narrative and workflow assistance around those scores, never to produce them.
Known limitations
AI-assisted summaries can occasionally misstate nuance, compress context in ways that read as more definitive than intended, or lag behind the latest source filing. AI output does not replace source filings, reviewer judgment, or the correction process — public evidence, methodology rules, and documented review decisions remain the basis for ArtMetrics trust materials.