Two auditors reviewed the same 32 companies for AI hiring bias. New York City's DCWP found 1 violation. State auditors found 17. That gap is the whole story of where AI hiring compliance sits in 2026. A company hiring in NYC, Chicago, Denver, Austin, London and California is now inside six overlapping AEDT regimes. Illinois HB 3773 and Texas TRAIGA went live January 1. California's FEHA automated-decision rules took force last October. Colorado's AI Act lands June 30. The EU AI Act treats recruitment as high-risk from August 2. Here's the trap most teams miss: a single "universal" bias audit passes none of them cleanly. NYC LL144 demands intersectional race-by-sex impact ratios. Illinois bans zip codes as proxies — but the EU AI Act's data-representativeness requirement often needs exactly the geographic coverage zip codes provide. Strip them out, fail Brussels. Keep them, fail Illinois. Texas rejects disparate impact as a theory entirely; California accepts both impact and intent. Six regimes. One hiring stack. No single audit shape fits all of them. And that's before the litigation. Mobley v. Workday established that an AI vendor can be liable as an "agent" when its tool filters candidates. Kistler v. Eightfold is testing whether AI hiring platforms are FCRA consumer reporting agencies — if it lands, every scored candidate is owed an adverse-action notice. The "our vendor says it isn't an AEDT" memo from 2024 is now a discovery exhibit. And the price isn't theoretical: LL144 runs up to $1,500 per day, per violation — over half a million a year on a single unaudited tool. If you run AEDTs across more than one jurisdiction, save this — the conflict map is the part nobody hands you. What's sitting in your AEDT inventory that nobody has audited yet? #AIGovernance #HRTech #AICompliance #AlgorithmicBias #EmploymentLaw