Provenance Record

Strong, well-structured breach News piece (709 words) with disciplined single-outlet attribution of every figure, date, and data-category list, and a careful 'What We Don't Know' on attribution and negotiations.

All 4 snapshots read from disk (HTTP 200). source-2 (SecurityWeek) confirms the 5,995,277 figure, the April 14 identification, the social-engineering vector, the government-issued-ID data category, the 24 months of credit monitoring, the ShinyHunters 8.7-million-record claim, and the prior-breach history (2019 hack, 2020 ransomware, March 2021). source-3 (BleepingComputer) confirms the 'An unauthorized actor used social engineering to deceive an employee…' notice quote, the April 10 initial access, the April 22 'first determined that the bad actor illegally copied' date, the 5,995,277 figure, and the 8.7-million-records / terabytes claim. source-0 (The Register) confirms the TransUnion 2-year monitoring, the 'In addition to the comprehensive security measures…' company quote, the Have I Been Pwned 8.7-million figure, the 'hinted at a breakdown in negotiations' framing, the verbatim ShinyHunters 'incredible patience' quote, and the declined-to-comment note. source-1 (The Record) confirms the 'We acted swiftly to block the unauthorized activity…' company quote, the driver's-license/passport data categories, the 'not publicly attributed' point, and the 'gained access to a limited portion of its IT environment … after compromising an employee account' framing.

Every figure, date, and quote is verbatim in the cited outlet. The cross-outlet discrepancies the bot flagged are handled correctly: the April 10 / April 14 / April 22 dates are each attributed to the single outlet that carries them, and the two differing data-category lists (SecurityWeek vs The Record) are attributed separately with no spliced/compound citation. No fabricated specifics.

High-quality, fully-verified breach submission with model cross-source-discrepancy handling. APPROVE — the single automated warning is a verified false positive.