Methodology · v3.2

FEUHistorique FTC (US)

FTC enforcement actions against {legal_name} in last 5y. NEGATIVE-SIGNAL (higher = cleaner record). CRITICAL OUTPUT RULE: "no FTC action found" is a GOOD outcome (clean record), NOT missing data. If after searching ftc.gov you find NO FTC-specific action in 5y, you MUST return enumerated_events with events=[] (empty array) → server scores 95 (clean). NEVER return type="absent" — absent wrongly maps to insufficient and discards a clean-record signal. EXTRACTION RULES: ONLY actions issued by the U.S. Federal Trade Commission (ftc.gov) — enforcement actions DB (ftc.gov/legal-library/browse/cases-proceedings) or press releases (ftc.gov/news-events/news/press-releases). STRICT EXCLUSIONS: exclude CFPB (consumerfinance.gov), SEC, DOJ, state AGs, OFAC, EU regulators (those belong to compliance-history / regulatory-compliance-track). Each real event MUST cite an ftc.gov URL + date + severity. CLEAN-RECORD REASONING (MANDATORY to publish the clean 95): when you return events=[], your reasoning text MUST explicitly state ALL THREE of: (a) the official regulator source(s) you searched, named by hostname (e.g. "ftc.gov"); (b) the explicit time window searched, written as "last 5 years / 60 months"; (c) an explicit negation: "no enforcement action against this specific company was found in that window". A bare events=[] WITHOUT this three-part affirmation is treated as an INCOMPLETE search (not a verified clean record): the server bails it for retry and it will NOT score 95. Server applies event_count_inverse.