Texas AG Race Could Bring New Scrutiny of Workplace AI

Texas businesses using artificial intelligence to set compensation or select employees for layoffs could face greater scrutiny if Democratic attorney general nominee Nathan Johnson wins in November.

The state senator’s proposal targets two gaps in Texas’ existing AI framework: tools used to manipulate wage bargaining and the absence of required disclosures when AI helps determine which workers or positions are eliminated. According to an August 26 Fisher Phillips analysis, his election could change enforcement priorities even if broader legislative changes prove elusive.

In his campaign’s “Catching Up To Yesterday” plan, Johnson promises an audit of consumer protection, competition and employment laws during his first 30 days in office, followed by work with legal teams and lawmakers to address identified shortcomings.

“The laws designed to protect Texans must keep pace with advances in technology and marketing innovation,” the campaign document says. Johnson also proposes twice-yearly reports from the attorney general’s Consumer Protection Division identifying technological developments that create vulnerabilities for consumers and employees.

For employers, the potential shift would extend scrutiny beyond discriminatory hiring and firing to how algorithms influence compensation negotiations and workforce reductions. A disclosure requirement could force companies to explain when automated tools contributed to termination decisions, including decisions to eliminate entire positions. Restrictions on wage manipulation could require closer examination of compensation software and vendor practices.

Those remain policy objectives, however, rather than enacted requirements or detailed proposed rules. The campaign has not specified what conduct would constitute prohibited wage manipulation or precisely what employers would need to disclose.

Texas already regulates certain AI uses through the Texas Responsible Artificial Intelligence Governance Act, effective Jan. 1, 2026. TRAIGA prohibits developing or deploying AI with the intent to unlawfully discriminate against protected classes; disparate impact alone does not establish that intent. It also restricts certain harmful manipulation, government social scoring and biometric uses. Enforcement rests exclusively with the attorney general, with no private right of action. (Texas attorney general guidance.)

That enforcement structure makes the election consequential, but does not give the winner unrestricted authority to rewrite the statute. Fisher Phillips cautions that Republicans are strongly expected to retain both legislative chambers and that attempts to stretch TRAIGA beyond its limited scope would probably draw litigation from business groups.

Johnson nevertheless has a plausible electoral opening. The July 15–17 Texas Public Opinion Research pollcited by Fisher Phillips put Republican Mayes Middleton ahead 39% to 38%, with 18% undecided and a margin of error of 3.4 percentage points. An August 7 Texas Pulse poll subsequently showed Middleton leading 46% to 44%. The findings suggest a competitive race, not an assured Democratic breakthrough.

Businesses need not wait for the election to examine their exposure. Fisher Phillips recommends documenting where AI affects hiring, promotion, discipline and termination, along with the human review applied at each stage. Reviewers should have actual authority to override automated recommendations.

Employers should also ask vendors how their systems were tested for bias and what safeguards address manipulative design, particularly in compensation and bargaining tools.

The immediate business issue is preparedness: knowing which systems influence employment decisions, preserving evidence of meaningful oversight and distinguishing existing legal duties from campaign proposals. A Johnson victory could intensify enforcement quickly, while broader workplace AI obligations would face a more uncertain legislative path.

The post Texas AG Race Could Bring New Scrutiny of Workplace AI appeared first on PYMNTS.com.