For years, children in the United States learned how to read and write with phonics. Phonics teaches the entire alphabet and how to properly pronounce each letter or group of letters. Children learn the many pronunciations that vowels have, how “Y” is a consonant but often acts as a vowel, and other basics of the language. Over time, phonics was replaced in many states with the whole language and balanced literacy models. These teaching methods were lauded at first, but faced growing great criticism as reading scores declined. Now, AI tools are showing promising results with teaching phonics, bringing this traditional method back into the spotlight.

The Phonics Alternatives: Whole Language and Balanced Literacy 

What exactly are these models? 

The whole language model prioritizes learning how to read by the natural association of certain objects or pictures. These models are designed to have students rely on using context clues and inferences to help them understand what word they are looking at. It believes that reading is a skill that happens naturally, and therefore, learning how to sound each letter out is not as important in learning how to pronounce words and how they go together. On the other hand, while balanced literacy requires some knowledge of phonics, it is not highly prioritized in this model. Simply put, balanced literacy combines elements of traditional phonics and the whole language method. 

Neither of these styles are particularly new. The whole language model dates back to the 1800s when the “father of American education” and politician, Horace Mann, declared that teaching children to sound out words would distract from the true meaning of them. By the 1950s, many schools were beginning to use the whole language model to teach their children how to read by prioritizing the meaning of words over pronunciation. Balanced literacy started in the 1990s, as teachers wanted to have the students continue reading books but with a sprinkle of learning phonics. 

As these methods grew in popularity in the early to mid-2000s, reading scores began to decline. In only a short, five-year period, it was clear that balanced literacy was not working for students. A 2024 study showed that twelfth-graders had a lower average reading score in 2024 than those in 2019. While the changes that occurred within this five-year timeline are relatively modest, it is a microcosm of a greater literacy problem within the United States. 

This steady decline is at the root of the “reading wars,” the clash between traditional phonics-based instruction and the whole language or balanced literacy approach. While this debate is far from settled, AI tools are now shifting how traditional phonics can be delivered in the classroom. This pivot is supported by research from nearly two decades ago. 

The National Reading Panel was created in 1997 by the U.S. Department of Education. In 2000, they provided an in-depth assessment on reading instruction that concluded that a systematic, explicit, and direct approach to learning phonics is essential for early reading understanding. Despite the evidence, alternative methods have continued to prevail in schools. 

AI in the classroom? 

Can AI now be considered as a feasible alternative (or addition) to relearning phonics across the classrooms that replaced it with whole language and balanced literacy? 

There are multiple AI tools that are being introduced for students to use. One of the most popular is Amira Learning, voice-recognition software that allows the student to read stories aloud while providing an interactive learning experience. Teacher’s College at Columbia University reported that students who used Amira gained more fluency, comprehension, and a broader vocabulary within only eight weeks. 

Amira is not only aiding in learning phonics, but it is doing so at an exceptionally high rate. There are an average of 36 weeks per school year; Amira users have shown more than 45 weeks of growth in that time. 

How does this actually help students better learn phonics? There are three major factors that make Amira a feasible tool. The first is speech recognition. Reading and speaking aloud to a teacher and other students was one of the primary ways traditional phonics was taught. While there is no physical instructor available, Amira mimicking the reading aloud model is proof that the traditional method of learning phonics worked and that it can be applied to other methods of teaching, such as with AI. Interactive learning plays a major role in students learning phonics. When a student mispronounces a word, Amira immediately intervenes and corrects the phoneme error. Finally, there is data tracking, which makes it easy for the parents, teacher, and / or student to see where they are having trouble and where they are successful. 

What are the benefits and drawbacks?

The benefits of using AI aren’t always so black and white. Education Week supports Amira’s model of providing a personalized learning experience and feedback tailored to individual students’ needs. A Stanford University review shows that using these tools to learn versus using them to do their work improves the students’ learning experience. 

At the same time, experts express concern that these tools may offload too much work onto technology, taking away from a more robust learning experience. Susanna Loeb, professor at the Stanford University Graduate School of Education contributed to the review linked above. 

“Unless we are careful with chatbots, they can really take away from learning,” Loeb told Education Week. “Because it is so tempting to give over the learning to the chatbot to produce something.” 

Some schools are putting on the brakes on the new technology while they study its impact. A recent AI ban in New York City has placed major restrictions on generative AI in grades K-8 for the next year. “The tech industry wants us to believe that A.I.-powered early education is not only inevitable, but necessary,” said Mayor Zohran Mamdani, announcing the policy. “We do not see it that way.  Students will still be able to use a limited number of AI tools and will take “AI critical thinking” courses alongside them. 

Moving forward with AI 

Where does this leave children who still need to learn phonics efficiently? The studies have continuously shown that the traditional method is tried and true. Yet, these emerging AI methods are proving themselves to be effective while not overwhelming them and providing additional time for the student and parent to learn outside of the classroom. Combining the two, in a manner where the student is still doing the majority of the learning on their own, can help ensure that no student is left struggling as schools begin to reintroduce phonics into the classroom. By incorporating traditional frameworks with digital innovations like Amira, school districts can offer a path forward for an individualized practice that allows the teachers to return to the drawing board of phonics.