Texas AG nominee Nathan Johnson pitches a 30-day “AI audit” to rewrite state rules
The plan targets consumer, competition, and employment laws and would land recommendations at the Legislature within his first month in office.
Texas Democratic attorney general nominee Nathan Johnson is campaigning on a first-30-days “AI audit” that would review state consumer protection, fair competition, and employment laws for AI-era gaps. If elected, he says he would deliver a concrete package of legislative recommendations to the Texas Legislature within 30 days.
Key Takeaways
- Texas Democratic AG nominee Nathan Johnson is proposing a first-month “AI audit” that would produce legislative recommendations for consumer protection, fair competition, and employment laws.
- Johnson is framing AI-driven firing decisions as a disclosure gap, arguing Texas does not require employers to tell workers when AI is used to determine who gets fired.
- Texas already has an AI-specific statute in force, the Texas Responsible AI Governance Act, which took effect at the beginning of 2026 and is enforced by the attorney general.
- Johnson is contrasting his enforcement posture with incumbent Ken Paxton’s, while Paxton’s office did not respond to a request for comment on the criticism.
Johnson’s 30-Day “AI Audit” Pitch for the Texas AG Office
Nathan Johnson, the Democratic nominee for Texas attorney general, is running on a plan to conduct what he calls an “AI audit” of Texas law, with a hard deadline attached. The deliverable is not a lawsuit or a guidance memo. It is a specific list of legislative recommendations delivered to the Texas Legislature within his first 30 days in office, if he wins.
The scope Johnson laid out is broad by design. He wants the review to cover consumer protection law (rules against deceptive or unfair practices), fair competition law (rules meant to prevent anti-competitive conduct), and employment law. The pitch is that the attorney general’s office should not only enforce existing statutes against AI-related harms, but also push lawmakers to update the statutes themselves.
Johnson framed the rationale as speed. “Stuff that we thought the statutes were fine to handle three years ago might as well have been 3,000 years ago. The pace is changing so rapidly,” he said.
Where Johnson Says Texas Law Still Falls Short on AI
The “AI audit” concept, as described, is a structured gap analysis: take the laws the state already uses to police consumer harm, competition issues, and workplace practices, then identify where AI changes the fact pattern enough that the current language no longer bites. The output is meant to be a prioritized list of statutory updates, not a general statement that AI is risky.
Johnson’s most concrete example sits in employment. He argued Texas law does not require employers to disclose when they use AI to determine who gets fired. That is a narrow claim with a wide blast radius, because it points at a transparency requirement rather than a ban. If that framing becomes a legislative template, it can extend beyond termination decisions into other AI-mediated employment decisions where disclosure is cheap to mandate and hard to litigate after the fact.
He also put consumer protection and fair competition in the audit’s scope, which matters because those categories are where state attorneys general typically have the most flexible enforcement tools. The mechanism here is not new authority on day one. It is the attempt to convert existing enforcement lanes into AI-specific statutory language quickly, so future cases do not depend on stretching older doctrines to fit new systems.
Texas’ Existing AI Enforcement Backdrop and the Paxton Contrast
Texas is not starting from zero on AI-specific law. The Texas Responsible AI Governance Act took effect at the beginning of 2026 and prohibits AI designed with specific harms, including discrimination or child exploitation. Enforcement sits with the Texas attorney general, which means the office already has a defined statutory lane for AI-related actions.
Johnson’s proposal would sit on top of that framework. The audit is positioned as a way to expand the state’s toolkit beyond prohibitions aimed at clearly harmful systems and into the gray zone where AI changes how decisions are made, documented, and explained.
Johnson also made the enforcement posture part of the campaign contrast with incumbent attorney general Ken Paxton. “What might distinguish me more from Ken Paxton is I'm not going to be selective in my enforcement,” Johnson said. He went further, alleging: “Paxton has gone after a lot of tech giants, but he leaves alone anyone affiliated with Elon Musk. I'm not going to apply the law to everybody except to my political friends and allies.”
The packet does not independently substantiate that selectivity claim. What is confirmed is that Paxton has brought lawsuits against tech companies including Meta and Character.AI, and that Paxton’s office did not respond to a request for comment on Johnson’s criticism.
Signals for Crypto and AI-Heavy Firms Operating in Texas
For traders and operators, the near-term signal is not a single enforcement action. It is the speed promise. A first-30-days recommendation package is a way to compress the timeline between “AI is changing things” and “here is the bill text,” assuming Johnson wins and lawmakers engage.
Four practical markers matter more than the campaign rhetoric.
First, election probability. Polling and any updated odds that move the likelihood Johnson takes office are the gating factor, because the audit is contingent on winning.
Second, specificity. Any publication of Johnson’s recommendation list, or even draft priorities, will matter most where it names concrete disclosure duties or standards in employment, consumer protection, or competition. A general call for “responsible AI” is cheap. A requirement to disclose AI use in employment decisions is operational.
Third, legislative uptake. Bill prefilings or committee agendas that track the audit’s themes would be the first sign the proposal is becoming a legislative program rather than a campaign line. That is where compliance expectations start to harden.
Fourth, baseline enforcement under the Texas Responsible AI Governance Act. New guidance or public actions clarifying how the statute is applied would set the floor for what the next attorney general inherits, and it would shape how much incremental risk an “AI audit” could add.
My Read: This Is a State-Level Compliance-Risk Story, Not a Token Catalyst—Yet
The part that decides this is not Johnson’s critique of Paxton. It is the promised artifact: a list of legislative recommendations delivered on a 30-day clock, which is a mechanism for turning AI anxiety into bill language fast if the election breaks his way.
The threshold that matters is whether the audit produces a narrow, enforceable transparency template, starting with employment disclosure around AI-driven firing decisions, and whether the Legislature signals it is willing to pick that up. If that happens, Texas moves from “AG enforces a harms-based AI statute” to “Texas sets new default disclosure expectations for AI-using firms,” and that is when the risk becomes operational rather than rhetorical.
