AI-Governance Claims, Checked Against Primary Sources
Why this report exists
Most claims about AI never get independently checked. They get repeated — and in a field moving this fast, a "fact" that was true six months ago can be repealed, deferred, or never enacted at all. The decisions built on it inherit the error, silently.
Second in a standing series. We take claims circulating widely in the AI-governance market and check each against the primary source — the statute, the regulator, the official record. Nothing here is a hot take; every verdict is sourced. Two disciplines govern the report:
- We verify claims, not claimants. No company is named. We are not grading anyone's product or character — only whether a statement holds against the record.
- Precision over posture. The goal is not to "debunk." It is to state the exact position, including the nuance. An overcorrection is just a new error.
Verdict scale · VERIFIED / PARTIALLY VERIFIED / NOT VERIFIED
A 10-year moratorium on state enforcement of AI laws passed the U.S. House inside the budget reconciliation bill (the "One Big Beautiful Bill Act") on May 22, 2025. The U.S. Senate stripped it by a 99–1 vote on July 1, 2025 — a revised version conditioning it on BEAD broadband funding, and a Blackburn–Cruz compromise, both collapsed. The President signed the bill on July 4, 2025 with no moratorium in it. States never lost the authority to regulate AI.
"There is a federal moratorium" is false. But "the idea is dead" overstates it — federal preemption proposals continue to circulate. As of mid-2026, no federal moratorium exists, and state AI laws remain enforceable.
TRAIGA (HB 149) was signed in 2025 and takes effect January 1, 2026. It reaches entities that develop or deploy AI in Texas or serve Texas residents, and prohibits specific uses — behavioral manipulation, unlawful discrimination, unlawful deepfakes / CSAM, and infringement of constitutional rights. It adds disclosure, social-scoring and biometric restrictions for government entities, is enforced by the Texas Attorney General (with a cure period), and creates a regulatory sandbox and an AI advisory council.
The effective date and existence are verified. But "a sweeping mandate on every business" is overstated: several private-sector prohibitions are intent-based, and the heaviest obligations fall on government use. Real and imminent — narrower than the headlines.
SB 1120 — the Physicians Make Decisions Act — took effect January 1, 2025, not 2026. It does not ban AI in utilization review. It requires that an AI or algorithmic tool not be the sole basis for a medical-necessity denial; that a licensed clinician make the final medical-necessity determination; that decisions rest on the individual patient's information, not solely group datasets; with transparency and oversight by the California DMHC and Department of Insurance.
The core — a human, not an algorithm, makes the final medical-necessity call — is real and holds. But "bans AI" overstates it (AI may still assist), and the effective date is January 1, 2025, already in force. A 2026 roadmap citing it as an upcoming ban is wrong on both the date and the scope.
The pattern — and why it matters
Three widely-repeated claims: one flatly false (the moratorium was removed), one true (Texas, on Jan 1, 2026), one true-but-garbled (California — right principle, wrong year and wrong scope). Each is shaping decisions right now, and each is a 30-second check against the public record away from being right.
That is the quiet risk under the loud AI-governance conversation. A compliance roadmap, a board memo, a market thesis, or an AI system's own "we're compliant" dashboard is only as current as the last time someone independent checked its inputs against the source. Self-report and repetition are not verification. This report is a small, public demonstration of the difference.
What this report is — and isn't
- It is an independent check of public claims against primary sources, issued at no charge.
- It is not legal advice, and not a compliance opinion for any organization. Confirm your own obligations with counsel and the current statute text.
- It names no company and evaluates no product. It grades statements, not people.
MarginSignal OS — independent verification of what deployed AI, and the claims around it, actually do.
Corrections to any finding in this report are welcome and will be issued publicly. That is the point.