OKLAHOMA CITY (OBV) — Oklahoma is trying to do two things at once on artificial intelligence: push adoption to unlock productivity in a tight labor market while lawmakers advance a first wave of guardrails aimed at schools, minors, government transparency, and synthetic media harms.
According to polling of more than 350 likely voters in Oklahoma conducted for OBV in March 2026, AI tools are already a routine part of life for a meaningful share of the electorate, even as voters prioritize “keeping humans in charge” and protecting children from harmful chatbots.
At the Capitol, multiple AI-related measures are moving through the second regular session of the 60th Legislature, with bills spanning K-12 policy, minors’ access to “companion” chatbots, state-agency limits and deepfake protections. The session calendar adds urgency as crossover bills work their way through committees and floor action.
AI is already in the bloodstream — and not just in tech circles
According to the poll, over 50% of respondents utilize AI tools throughout their personal and professional lives. 19.9% said they personally use artificial intelligence tools such as ChatGPT, Gemini, Copilot, or other AI assistants daily. Another 10.2% said they use them a few times a week, and 19.3% said a few times a month.
That uptake is one reason AI policy is no longer a speculative conversation. Legislators are being asked to define rules for tools many constituents already use at work and at home—and that school systems are beginning to adopt in structured ways. At the same time, the public conversation continues to split between competitiveness and caution: how quickly to adopt, and what guardrails should come first.
What voters say they want: humans in charge, child protections, and transparency
When likely voters were asked to choose the top safeguard, “keeping humans in charge of important decisions” ranked first—notably ahead of protecting children from harmful AI chatbots. Preventing misuse of personal data followed, and disclosure when government uses AI-generated content was also included among voters’ priorities.
Those preferences line up with the dominant policy lanes moving at the Capitol: bills that emphasize human accountability and oversight, proposals that restrict certain chatbot designs for minors, and measures intended to deter deception and identity misuse through synthetic media.
The economic frame: “win the AI race” in a labor-constrained state
Business leaders arguing for faster adoption have emphasized that AI is not just a technology story, but a capacity story—one that intersects with Oklahoma’s labor market constraints and long-term competitiveness.
Hart Brown, president of AI and transformation at Saxum, has framed Oklahoma as being near an inflection point that could reshape the state’s economic trajectory if leaders move quickly. He has tied the argument directly to workforce limits.
“We are labor constrained both in the United States and here in Oklahoma with a really low unemployment rate,” Brown said. “It becomes very difficult for CEOs and owners of businesses to grow because we just can’t find the people.”
Brown has also argued the policy environment could become a competitive advantage if Oklahoma provides clarity early. “Some state will become the next Delaware by writing AI legislation correctly to protect the companies that are going through this process and to protect the consumers at the same time,” he said. “Every company on the planet will want to domicile in that state.”
But he has also pointed to infrastructure constraints that will shape where AI-heavy investment can land, particularly energy and water demands associated with data-center expansion. “The increases in energy are causing a bit of a concern because we haven’t had to build infrastructure that fast,” he said. “The increases in potential water use is a concern as well.”
That tension—move quickly, but build smart—is also reflected in the way many lawmakers describe their own approach. Brown said legislators are interested but wary of getting it wrong. “I need to learn more because I’m hesitant to write something that may in fact be wrong,” he said. “I need to understand the North Star, and I need to understand the technology. Let’s get the language right.”
The policy landscape: where the major AI bills are right now
Below is a “where we are” snapshot of the most visible AI measures advancing this session, organized by theme.
1) Schools: SB 1734 advances a statewide framework for classroom AI
One of the most defined proposals is Senate Bill 1734 by Sen. Ally Seifried, R-Claremore. The measure creates the Oklahoma Responsible Technology in Schools Act and establishes guardrails for AI use in public school districts.
The bill’s framing centers on educator control and transparency to families. Under the approach, student-facing tools are intended to be age-appropriate and used under the guidance of a classroom educator, with districts notifying parents when AI is used in the classroom and providing an opt-out option.
The Oklahoma Senate passed SB 1734 unanimously and it has moved to the House for consideration.
Polling context: The survey tested support for classroom AI in two ways—AI as a tool to assist teachers and AI as a tool students may use under supervision. The results show support is meaningful but not universal.
For teacher-assist uses (lesson planning, grading, administrative tasks, with teachers responsible for final decisions), the results show 10.6% strongly support and 17.0% somewhat support, while 21.7% somewhat oppose and 6.7% strongly oppose.
For student use in class (approved by the school, supervised by a teacher, and parents informed), the results show 15.6% strongly support and 23.3% somewhat support, while 21.3% somewhat oppose and 6.8% strongly oppose.
Together, those results suggest many voters will accept AI in schools when it is structured—supervised, disclosed and educator-directed—rather than open-ended.
2) Minors and “AI companions”: parallel efforts in House and Senate
A second lane of bills targets “companionship” and human-like relationship in AI systems accessible to minors—an area lawmakers have discussed as a child-safety issue rather than simply a content or privacy question.
House Bill 3544 by Rep. Cody Maynard, R-Durant, focuses on AI chatbots and “companions” in relation to minors. The measure has advanced through the House and crossed to the Senate, where it has been referred to the Senate’s Technology and Telecommunications committee.
In a separate Senate track, Senate Bill 1521 by Sen. Warren Hamilton has focused on safeguards for minors interacting with certain conversational AI systems. The bill has been passed through the Senate, with discussion focused on the need to encourage innovation and protect children.
Polling context: The poll asked about limits on AI chatbots “designed to imitate human relationships” and accessible to minors. The results show 20.0% strongly support and 12.1% somewhat support limits, while 24.8% somewhat oppose and 5.2% strongly oppose.
That split suggests the minors/companionship lane is more contested than general “humans in charge” guardrails, even as child protection ranks high among voters’ stated priorities.
3) State government guardrails: HB 3545 outlines “high-risk” limits and oversight
A third policy lane targets how state government itself can deploy AI. House Bill 3545 by Maynard provides definitions and establishes prohibited uses of artificial intelligence by state agencies, while restricting other uses and implementing reporting requirements.
The language outlines multiple prohibited categories, including cognitive behavioral manipulation, unlawful discrimination through automated classification, certain uses of real-time remote biometric identification for public surveillance, and deceptive deepfakes used maliciously. House messaging around the bill has emphasized human oversight for high-risk decisions and transparency about how AI is being used by government.
Polling context: The strongest signal from voters in the survey is consistent with that approach: “keeping humans in charge of important decisions” was the top safeguard choice. The poll also tested whether state agencies should be required to disclose when AI is used to generate public-facing content or inform important decisions. The results show 21.6% strongly support and 7.5% somewhat support that requirement.
Those results do not settle the policy question, but they reinforce that “human accountability and transparency” is a politically durable theme for AI rules in state government.
4) Deepfakes and identity misuse: HB 3299 moves to penalize harmful synthetic media
Beyond classroom rules and minors’ access, lawmakers are also focusing on synthetic media harms—particularly deepfakes that replicate a person’s likeness, voice or identity.
House Bill 3299 by Rep. Neil Hays, R-Checotah, would make it unlawful to create and distribute certain digitized or synthetic media depicting another person’s name, image, voice or likeness without written consent and with intent to cause emotional, financial, reputational or physical harm. The measure also addresses disclosure requirements for synthetic media used in political advertising and did not advance through the House process.
For business readers, the relevance extends beyond elections. Synthetic identity misuse is tied to reputational risk, consumer trust, and fraud exposure—issues that can affect companies and individuals alike as synthetic media becomes easier to produce and harder to detect.
5) First-principles definitions: HB 3546 and the “personhood” line
Another Maynard measure, House Bill 3546, addresses a foundational legal question: whether AI systems could ever be treated as “persons” under Oklahoma law. The proposal is intended to clarify that AI systems and algorithms may not be granted legal personhood, reinforcing that rights and accountability remain with human beings rather than machine systems.
While less operational than policy on school or minors’ access, the “personhood” line fits within the same accountability frame voters elevated in the safeguard question: humans should remain responsible for the decisions and outcomes AI systems influence.
6) The data layer under everything: Oklahoma’s new privacy law
Even where bills are not labeled “AI,” many of the sharpest AI policy conflicts run through data: what gets collected, how it is processed and what consumers can control. That backdrop has shifted with Oklahoma’s newly signed data privacy law, Senate Bill 546, which establishes consumer rights and rules for businesses that process personal data.
As AI adoption accelerates, privacy compliance becomes part of the operating environment for businesses using AI tools and data-driven systems—especially when personal or sensitive information is involved.
What happens next: deadlines, committee choke points, and “getting the language right”
With the session calendar moving toward major end-of-session deadlines, the pace of movement will matter as much as the policy substance. The Senate calendar sets May 7 as the third-reading deadline in the opposite house and May 29 as the constitutional mandate for sine die adjournment. An early budget agreement could allow lawmakers to adjourn ahead of that date.
Brown’s “write it right” warning captures the broader legislative posture: lawmakers are under pressure to act, but also to avoid rules that become obsolete or counterproductive as the technology changes. “Let’s get the language right,” he said.
AI legislation tracker (60th Legislature, 2nd Session) – current status
- SB 1734 – Sen. Ally Seifried – Engrossed to House; House First Reading (03/24/2026).
- HB 3544 – Rep. Cody Maynard – In Senate; Second Reading referred to Senate Technology & Telecommunications (04/01/2026).
- SB 1521 – Sen. Warren Hamilton – In House; last posted action: Do Pass from Government Modernization and Technology (04/06/2026).
- HB 3545 – Rep. Cody Maynard – In Senate; Second Reading referred to Senate Technology & Telecommunications (04/01/2026).
- HB 3299 – Rep. Neil Hays – House measure; last posted action: Do Pass (as amended) from House Judiciary & Public Safety Oversight Committee (02/26/2026).
- HB 3546 – Rep. Cody Maynard – In Senate; Second Reading referred to Senate Technology & Telecommunications (04/01/2026).
- SB 546 – Sen. Brent Howard – Approved by Governor (03/20/2026) (enacted; no longer pending).
