The cheapest public records request is the one nobody has to file.
That is roughly the argument Rep. Cody Vasut makes when asked how Texas should handle a problem arriving faster than the law anticipated: artificial intelligence has made it trivial to generate public information requests by the hundreds, and the government employees who must answer them are paid by the same taxpayers footing the bill for the search.
Vasut’s answer is not to make requesting harder. It is to make requesting unnecessary.
“Just put more stuff online that can be accessed without a PIA, so that you’re not having to deal with as many PIAs,” he says, using the shorthand for the Texas Public Information Act. Many public documents that are the subject to TPIA requests — budget documents, financial statements, financial records — those can simply be published automatically.
He is quick with the caveats, which are the caveats of a practicing attorney. Don’t post the city secretary’s signature into the ether, because the check could be forged. Don’t post bank account numbers, routing numbers or SWIFT codes, because someone will empty the account. Redact those.
“And then just, you know, put the checkbook out there,” he says, “and maximize the amount of stuff that you’re putting online to begin with, so that people don’t have to do as many requests.”
Then the principle, stated plainly: “I am not an IT expert. But I just generally think, as a basic policy, the more information online and accessible — to the extent it doesn’t invade individuals’ privacy or financial data — the better.”
It is a proposition with an unusual political profile: it costs local governments less over time, and it gives the public more.
The chairman
Vasut chairs the House Select Committee on Government Oversight, which Speaker Dustin Burrows created in March and which is charged with a deliberately broad question — whether government in Texas is working the way it is supposed to.
He has represented House District 25, covering western and southern Brazoria County, since 2021, and is heading into his fourth session. He practices law when the Legislature isn’t sitting. Before Austin, he served four years on the Angleton City Council, which turns out to matter a great deal to how he reads these issues.
Asked why the speaker picked him, Vasut demurs, then offers the practical answer: “I like to say I run a good committee meeting.”
The committee’s docket is genuinely miscellaneous — the Texas Tort Claims Act and the Texas Regulatory Consistency Act; public information and open meetings; school district education foundations; public money routed to nongovernmental organizations; prosecutorial discretion; and, improbably, the question of whether Texas could annex several counties in eastern New Mexico. Vasut insists the through line is real.
“They all come back to whether government is working effectively,” he says.
As transparent as possible
The Texas Public Information Act traces to a specific failure. In 1971 and 1972, the Sharpstown stock-fraud scandal reached the highest levels of state government, and a bipartisan bloc of thirty House members — the Dirty Thirty — kept the issue alive against a speaker who wanted it buried. The reforms that followed in 1973 opened most government records to public inspection, expanded open meetings requirements, and imposed disclosure rules on lobbyists and officeholders.
The statute’s stated purpose is expansive. Vasut’s interest is in the gap between that purpose and daily practice, and he frames it as two questions.
The first: are local governments actually following the law? He is careful not to prejudge. “I can’t answer that sitting here right now, because I don’t know what every local government has done in response to every request ever.”
The second is the more interesting one. The Act promises transparency “as possible” — and some things genuinely shouldn’t be disclosed. Trade secrets belonging to private companies doing business with government. Confidential terms in an economic development negotiation, where revealing your position to a competitor changes the deal. Law enforcement records tied to an open criminal investigation.
However, Vasut says, “those exceptions should not swallow the rule. And that’s the question.”
To answer it, his committee structured its hearing to take all three perspectives in sequence: the entities that must respond to requests, on compliance and process; the requesters, on whether they get what they ask for and whether the fees charged for staff search time are reasonable; and law enforcement, on the investigative exception.
The machine problem
Which brings the discussion back to AI, and to a genuine tension rather than a talking point.
The volume problem is documented. Governments across the country report being swamped by automated requests; the Pennsylvania Office of Open Records has warned that AI-generated filings have arrived containing fabricated legal citations, inaccurate case-law summaries and quotations from court decisions that do not exist — forcing staff to sort real arguments from hallucinated ones. Whole operations now exist to file records requests at scale and resell what comes back.
Vasut’s first observation is that the technology cuts both ways. “Both sides can access AI,” he says — to draft a request and to respond to one.
But he does not dismiss the cost. “Anytime you task local government with doing something, it costs money,” he says. “Government doesn’t act as — it’s not a magical being. It works through human beings.” That is why the Act contains a cost-recovery mechanism in the first place: so that the public gets its information without the search itself becoming an unbudgeted expense. “We want it to be net neutral.”
The trouble is that machine-generated requests can be sprawling and overbroad in ways a human drafter would not attempt, which pushes costs up and, if the fees rise in response, prices ordinary citizens out of a right that belongs to them.
Publishing first breaks the loop. It also creates a clean answer to a request: it’s already online. Vasut takes the thought one step further, into records retention — if a document was posted for the full period a government was required to keep it, and nobody asked, “you could have accessed it and you didn’t.”
Texas already has the scaffolding for this. Since 2016, the Comptroller’s Transparency Stars program has recognized cities, counties, school districts, community colleges and special districts that post financial documents, contracts and procurement records, debt obligations, pension data and open government materials online in usable formats. It is entirely voluntary. The open question for the 90th Legislature is whether some portion of it should be standard practice.
Where the money goes
The same instinct runs through the committee’s other charges, all of which are ultimately about following public dollars after they leave the treasury.
One is the use of nongovernmental organizations to perform what are essentially governmental functions — running a homeless services program, for instance, as distinct from ordinary vendors like the company that sells a city its patrol cars. Vasut, the former councilman, is wary of the state micromanaging local decisions.
“I’m a huge local control over local issues guy,” he says. “The state neither has the resources of time nor efficient ability to micromanage local government.”
But he draws a distinction between regulating the spending and disclosing it, and he thinks the second obligation holds regardless of where you land on the first. A resident in the political minority in their own city has limited recourse when they believe money is being wasted — they can testify, and they can vote — and neither works if they cannot see the transaction.
“Even if the state doesn’t have any regulatory skin in the game, you, the taxpayer, deserve to know where every single penny of that is going and to have it come into the light so that you can evaluate it,” he says. He wants the flow traceable in both directions: which organizations receive the money, and how those organizations are connected, directly or indirectly, to the officials voting to send it.
A related charge examines the education foundations attached to school districts across Texas — the nonprofits that fund classroom grants, equipment and supplies beyond what state formulas and local taxes cover. Vasut’s own district has one in Angleton, and he speaks warmly about the teacher grant awards it makes. His concern is narrow and structural: these entities were built to receive private charitable donations, not to serve as investment vehicles or holding accounts for public funds.
“There’s an open question about what the law is,” he says — whether public money moving into such a foundation should be permitted at all, and if so, under what conditions.
Discretion and its limits
The committee will also examine prosecutorial discretion, an area where Texas has already legislated. House Bill 17, passed in 2023 and authored by Rep. David Cook, expanded the definition of official misconduct to include adopting a policy of categorically refusing to prosecute certain offenses, and created a removal process any six-month county resident can initiate before a judge from a neighboring county. Gov. Greg Abbott has since floated going further, with expanded impeachment authority and a state prosecutor able to step in.
Vasut is notably careful here. He declines to say whether any particular district attorney in Texas has crossed the line, framing the inquiry as academic rather than targeted. He also flags the practical difficulty with the existing remedy: removal requires residents to sign a petition and go on the record against the official who prosecutes crimes in their county, and “some people may fear that they’re a target.”
His most useful contribution is a distinction the debate usually flattens. A district attorney who declines to prosecute a category of crime may be making an ideological choice — or may simply be out of assistant prosecutors.
“Resources versus willpower,” as he puts it. He can readily imagine a rural DA saying, “I don’t have the resources to prosecute every crime,” and suggests a state prosecutor’s most valuable function might be helping those offices rather than overriding them. Even in a county the size of his own, he notes, a finite number of assistant DAs is why so many cases end in pleas rather than trials.
The New Mexico question
Then there is the charge that has drawn the most attention and involves the least precedent.
Two state legislators in eastern New Mexico have proposed letting voters in contiguous counties petition to leave the state; the counties in question — Lea, Roosevelt and Eddy among them — sit atop the Permian Basin and, by their representatives’ account, generate an outsized share of New Mexico’s general fund while getting energy policy they don’t want. Burrows directed the House to study what Texas would do if asked.
Vasut treats it as a legal question first. One path runs through both legislatures agreeing and Congress ratifying the transfer. Another could run through the U.S. Supreme Court’s original jurisdiction over boundary disputes. He notes Texas has done this before — the state claimed Greer County for decades during a dispute over which fork of the Red River marked the line, and he has eaten lunch under a map at an Angleton Mexican restaurant that still shows it.
“It’s already looked different before in the past,” he says.
What success looks like
Vasut is realistic about what a legislative committee can accomplish. It cannot fire anyone. It has no enforcement power. What it has is a hearing room and a report.
“We are here to raise awareness so that either A, laws can be changed; B, laws can be enforced; or C, the voters can hold people accountable,” he says.
His council years leave him unwilling to caricature local officials — he knows budgets have to cover police raises and street repairs, and he thinks legislators who “flippantly say all local governments spend too much money” are talking about places they’ve never examined. But he also holds a view that shapes everything else on his docket: every dollar government spends “comes from a taxpayer through force,” and that fact imposes a duty to show the work.
“Make the best choice for the taxpayer,” he says, “not the level of government.”