Why Texas Rejected An AI-Driven Charter School Then Approved It For Vouchers: A Parent’s Breakdown
The rain streaked the coffee shop windows last Saturday, as my 7th grader, Jax, colored a T-Rex on a napkin and a group of us leaned in around the table, comparing school options for next fall. I’d had a crumpled brochure for a new, mostly online AI-driven school tucked in my purse for three months, so when my friend Sarah mentioned the latest news, I pulled it out, the edge already stained with a coffee ring. “Wait, they said no to the charter, right?” she said, wiping rain off her jacket. “How are they already approved for vouchers? That doesn’t add up.”
I’d been following the story because Jax has dyslexia, and our local public middle school hasn’t managed to fit his needs into their overcrowded schedule. He’s two grades ahead in math, but still struggles with decoding text, and I’ve spent the last year jumping between IEP meetings and searching for any option that doesn’t involve me quitting my job to homeschool him full time. Back in March, the Texas Charter School Commission rejected the school’s application, citing that it didn’t meet requirements for full-time certified staff and in-person accountability. Most instruction is tailored by AI, with teachers only doing weekly check-ins. The board said they couldn’t approve an untested model that didn’t meet state charter standards. Then, two weeks ago, the new education savings account (ESA) voucher program approved the school as an eligible provider. Suddenly, any family can use taxpayer money set aside for public education to pay tuition there, even though it didn’t pass the charter approval process.
Sarah’s 10-year-old has ADHD, and he’s been suspended twice this year for fidgeting and walking out of overcrowded classrooms. “I don’t care if it’s a charter or not,” she said, stirring her oat latte. “The public school isn’t working for him. If this lets him learn at his own pace without getting in trouble for being bored, why wouldn’t I take it?” Across the table, my friend Mia, who teaches 4th grade at the local elementary, frowned. “The rules for charters are there for a reason, right? If they couldn’t pass charter muster, what sort of oversight do they have now that they’re in the voucher program? No one’s checking if they’re actually teaching what they’re supposed to.”
That conversation got me thinking, because most of us aren’t policy experts, we’re just trying to figure out what this mess means for our own kids. I’ve spent the last couple weeks digging into the details of the program, testing their free trial, and talking to other parents on the waitlist, and I’ve landed on a few concrete steps any parent can take when checking out a new, untested program like this, instead of just taking marketing claims at face value.
First, don’t stop at the generic line that says “we employ certified teachers.” Ask specifically how many students each certified teacher is responsible for, and how often they interact one-on-one with your kid. When I emailed the admissions team for this AI school, I got a response within a day saying each teacher manages 150 students. That works out to less than two minutes of one-on-one time a week per kid, if that. Compare that to Jax’s current special ed case manager, who has 22 kids on her caseload. That’s a huge difference, and it’s never something programs put front and center on their homepage.
Second, ask what sort of accommodations they’ll provide for your kid’s IEP or 504 plan, and get the commitment in writing before you enroll. Charters are required by federal law to follow the Individuals with Disabilities Education Act, so they have to honor existing accommodations. But private ESA-approved providers don’t have the same requirement. I asked the AI school if they’d provide Jax with extended time on assessments, built-in text-to-speech for all reading assignments, and quarterly dyslexia progress checks, and the admissions rep said they “try to accommodate all needs” but wouldn’t put anything in writing before enrollment. That’s a red flag for me, because if it’s not written down, there’s nothing holding them accountable if they don’t follow through.
Third, ask to see a sample progress report and find out what sort of state oversight the program is required to follow. Charters have to administer the state STAAR test every year, publish their results, and undergo regular financial and academic audits. ESA providers only have to give some sort of annual assessment, but it doesn’t have to be aligned to state standards. When I asked this AI school for a sample progress report, they sent me a one-page sheet that only showed how many lessons the student completed, not what skills they actually mastered. That doesn’t tell me if Jax is actually catching up in reading, just that he clicked through the required number of modules.
I tried the free trial lesson with Jax last week, after dinner. The math lesson was great. It figured out within three problems that he already understood basic pre-algebra, skipped the whole month-long unit on one-step equations he’d already mastered, and moved him straight into two-step work that he’d been begging to learn. He lit up when he finished, said it was the first math lesson that didn’t feel like a waste of time. Then we did the reading lesson. The AI gave him a passage that was way above his decoding level, he got three questions wrong, and the algorithm just dropped him down to a lower level passage, instead of flagging that he struggles with letter recognition, not comprehension. No teacher stepped in to adjust the lesson to his dyslexia accommodations, no alert for a human to check in. That gap is obvious when you actually use the thing, not just read the website copy.
It’s easy to get wrapped up in the policy fight here, about whether vouchers are good or bad, or how AI fits into K-12 learning. But for most of us on the ground, this whole situation is just a symptom of a bigger problem: public schools can’t keep up with the demand for personalized learning, especially for neurodivergent kids. If the state says this model doesn’t meet the bar for a public charter, it’s odd that it’s good enough to take public money through vouchers. But that’s where we are right now, and more programs like this will pop up as the voucher program expands.
I still have the crumpled brochure in my purse, under Jax’s dinosaur napkin and a half-eaten granola bar. Jax keeps asking if we can sign up, because he hates getting called on to read out loud in his current class, and he’s bored out of his mind in math. I get it. But I’m still waiting for answers to the questions I asked, and I’m not in a hurry to jump into an unregulated program that won’t commit to his needs. We’ve got a few more weeks before enrollment closes, and right now, there’s no perfect choice.
Please indicate: Thinking In Educating » Why Texas Rejected An AI-Driven Charter School Then Approved It For Vouchers: A Parent’s Breakdown