Aldine ISD Turns to AI Reading Tools to Support Texas’s Emergent Bilingual Students

The music in a Worsham Elementary third-grade classroom on a weekday morning is not what you might anticipate. The sound of children reading aloud to a screen and the screen reading them back is layered beneath the typical rustle of paper and squeak of tiny chairs.

In Aldine ISD, which is located just north of Houston, the program known as Amira has practically taken over. Since the previous school year, the district has spent over half a million dollars on it—real money for a public system—and the leadership discusses it in the same manner as some districts discuss new buildings or buses.

In retrospect, this moment might appear to be the start of something greater. The official term for children learning English while still anchored in another language, typically Spanish, is emergent bilingual, and nearly half of Aldine’s approximately 56,000 students fall into this category. There are over a million of these students in Texas, but there aren’t nearly enough qualified teachers to meet their needs. Bilingual education has long been considered a “critical shortage area” by the Texas Education Agency. Administrators believe that recruiting alone won’t be enough to close that gap.

For what it’s worth, Amira is not ostentatious. The pupil receives a passage. The pupil reads it out loud. Depending on what the child needs at that precise moment, the program listens, highlights stumbles, modifies the difficulty, and provides real-time correction in either Spanish or English. The district’s chief academic officer, Faviola Cantú, refers to it as “just-in-time” support, and that term—which is taken from manufacturing—fits more neatly than it should. Instantaneous feedback is crucial, according to Elizabeth Gonzalez, a third-grade teacher at Worsham. Children listen to themselves, take note of the correction, and try again. She claims there has been a noticeable increase in their willingness to converse in English during regular class hours.

It’s difficult to ignore the fact that this isn’t limited to Aldine and is occurring throughout the Houston area. Prof. Jim Inc. has been hired by Houston ISD to assist with curriculum development. Teachers at Barbers Hill ISD use AI to create lesson plans and quizzes. Members of the Pearland ISD school board have awkwardly used ChatGPT to check library books for content that is forbidden by a recent state law. A large portion of the estimated $7 billion global ed-tech AI market is making its way into Texas suburbs and exurbs, which were not exactly thought to be leaders in classroom innovation a few years ago.

The part that causes skeptics to pause is the Aldine numbers. The district had the biggest increase in reading districtwide among large Texas districts, with a 7-point year-over-year increase, according to the 2025 STAAR results. Students received an extra six points on campuses using Coursemojo, a different AI teaching assistant platform that Aldine co-designed. There was an 8.5-point increase for emergent bilingual learners. 9.3 percent of students were eligible for free or reduced-price lunches. These gains are not the result of rounding errors. It used to take districts ten years of reform to generate these kinds of numbers.

It’s genuinely unclear if AI deserves the credit or shares it with diligent educators and an already-underway curriculum overhaul. Dr. Cantú takes care to avoid overselling. The co-CEO of Coursemojo, Dacia Toll, frequently emphasizes that the tool was created with teachers, not for them. Every conversation carries a subtle fear that a mistake could transform a helpful classroom assistant into the kind of tech-related error that schools have to spend years correcting. Concerns about data privacy persist. Concerns about equity persist. This also raises the question of what happens to a child who primarily learns to read by conversing with a machine.

Nevertheless, you get the impression that Aldine has discovered something that other districts will eventually imitate as you watch this play out. Perhaps not the particular software. Perhaps not the precise budget. However, the approach, which combines human instructors with tireless, patient software to fill in the gaps left by the system, seems likely to gain traction. The children reading aloud at Worsham are probably not considering whether that’s a quiet revolution or just a well-marketed pilot. All they’re doing is attempting to correctly pronounce the next word.