Generative AI Meets Biliteracy | AI for Bilingual Education

November 10, 2025
Luis Hernández

Category: AI in Education, Biliteracy, Literacy, Vocabulary Development

AI-Generated Audios

It was an honor to present at the Association of Literacy Educators and Researchers (ALER) Annual Conference 2025, held November 6–8 in Oklahoma City. This event brings together teachers, researchers, and innovators dedicated to advancing literacy across languages and communities.

I had the opportunity to present alongside my colleague Viviana Hall, co-author of the new bilingual children’s book ¿Hacia Dónde Va? / Where Are We Going?. Together, we explored how her book can serve as a foundation for AI-powered biliteracy instruction through a web application I developed as part of Enabling Learning’s research and instructional design work on generative AI.

Bringing Stories and Technology Together

Viviana’s story beautifully celebrates curiosity, friendship, and discovery, qualities that resonate deeply in bilingual classrooms. For our ALER session, we used the book as the core content for a Dual Language Lesson Planner, an AI-assisted tool designed to help teachers of emergent bilinguals quickly generate rich, standards-aligned lesson plans in English and Spanish.

During our live demonstration, participants saw how teachers can guide artificial intelligence through structured instructions and pedagogical frameworks. Instead of relying on generic AI prompts, we showed educators how to use retrieval-augmented generation (RAG) and curated datasets to ensure that every lesson generated by AI is accurate, culturally responsive, and instructionally sound.

Although this particular application was developed through coding (using the React JavaScript library to build the web interface and an API connection to Google Gemini for AI generation), it’s important to note that similar results can be achieved through no-code tools such as Custom GPTs from OpenAI or Google’s Gems. These platforms enable educators to design guided AI experiences without programming knowledge. If you’d like to explore this further, we cover these approaches in our free online course, “Boosting Productivity with AI,” where teachers learn how to create their own AI-powered assistants for lesson planning and daily work support.

How the System Works

Each time a teacher clicks the “Generate Activity” button in the web app, the system gathers and organizes several sources of information into a structured JSON request, a kind of digital blueprint or map of instructions that tells the AI exactly what to do and how to respond.

These data sources include:

The full text of the bilingual storybook in English and Spanish ensures that the AI grounds every activity in the authentic language, characters, and context of the story created by Viviana Hall.
Biliteracy teaching strategies based on research-backed principles for supporting emergent bilingual students, including scaffolding, language transfer, and cross-linguistic connections.
Question stems, sentence frames, and oral language scaffolds that help the AI design developmentally appropriate activities to promote comprehension, academic vocabulary, and speaking practice.
Explicit system instructions that define the AI’s role as an expert curriculum designer for second-grade bilingual classrooms. These instructions also incorporate the student’s language proficiency level (beginner, intermediate, or advanced) in either English or Spanish, allowing the AI to differentiate tasks accordingly.

This approach, known as Rich-Context Prompting, ensures that the AI receives a well-defined and pedagogically rich set of inputs before generating a lesson. By structuring the data this way, the system can produce consistent, high-quality lesson plans that remain aligned with the story’s content, the learner’s proficiency level, and both linguistic and academic goals.

After the activity has been created, a chat interface appears, allowing teachers to continue the conversation with the AI. Educators can ask follow-up questions about how to adapt the activity, request additional recommendations, or explore extensions and modifications for small groups or different proficiency levels. This feature transforms the app from a one-time generator into an interactive teaching assistant, supporting reflective, ongoing dialogue about instructional decisions.

A Human-Centered Approach to AI in Education

Throughout our presentation, we emphasized that AI works best when guided by educators. The goal isn’t automation, it’s amplification. When teachers define clear pedagogical frameworks, AI becomes a creative and efficient collaborator in designing meaningful learning experiences.

This is the essence of AI-driven instructional design: combining human expertise with technological precision to enhance, not replace, the art of teaching. The Dual Language Lesson Planner demonstrates how thoughtful prompt engineering and Rich-Context Prompting can align AI output with bilingual education principles, ensuring that every activity promotes comprehension, communication, and cultural awareness.

You can explore the interactive demo we presented at the Association of Literacy Educators and Researchers (ALER) conference here:

👉 https://apps.enablinglearning.com/bilingual-planner/

The demo allows users to select a reading standard, a proficiency level, and a language of instruction, and instantly generate a 15-minute bilingual activity based on Viviana Hall’s story. It’s a glimpse into how AI can support lesson design while preserving the human touch that defines great teaching.

We are deeply grateful to Dr. Araujo and the ALER leadership for their commitment to advancing literacy and for providing a space where educators and technologists can collaborate. Working alongside Viviana reminded me how powerful it is when storytelling, pedagogy, and innovation come together in the service of learners.

At Enabling Learning, we remain deeply involved in exploring AI for teacher preparation and support, ensuring that technology amplifies the educator’s role rather than replaces it. We recently discussed this vision in a book chapter published in a volume edited by Dr. Araujo, which explores AI-guided professional development for teachers in multilingual classrooms. You can read more about that work here:

👉 https://www.enablinglearning.com/a-new-chapter-ai-guided-teacher-support/

At Enabling Learning, we’ll continue developing AI-assisted tools that are transparent, research-based, and educator-led—helping teachers create engaging, culturally responsive learning experiences for all students.

Verónica Speaks Twelve Languages. The Hard Part Was the Rules.

Verónica, the AI tutor inside our courses and apps, now responds in twelve languages. Arabic, Telugu, Tamil, and Korean joined because districts and teacher preparation faculty kept describing the families they serve. Adding the languages was never the hard part, since these models were multilingual from the start. The hard part was the rules: which characters even count as letters, whether the page or the person gets to decide, and when a tutor should hold the target language instead of following a student into English. Here is what broke, what we changed, and why it matters.

A Certificate No One Can Verify. It Is Just a File.

Most professional certificates share one weakness: confirming they are real means contacting the issuer and waiting, so in practice it rarely happens. A PDF can be edited in a browser. A paper certificate proves only that someone owned a printer. This post makes the case for digital credentials for both learners and institutions, looks at what the 2026 hiring research actually shows, explains why course completion is the wrong trigger for issuing one, and describes transparently how our scored capstone gates work, why we chose the Proof of Knowledge platform, and what we would tell any institution starting from zero.

Claude for Teachers: Anthropic’s Free Program for Educators

On July 14, 2026, Anthropic announced Claude for Teachers, a program that gives verified K-12 educators in the United States a full year of free access to premium Claude capabilities. It includes a teaching skills library, lesson planning grounded in evidence-based curricula, class data analysis, and integrations with nine school tools. In this post, we break down what teachers actually get, the privacy commitments behind the program, and the honest caveats, including US-only eligibility. We also added a complete Claude for Teachers topic to our free course, Boosting Productivity with AI for Teachers.

Chalkboard reading Language Learning beside German, English, and Spanish dictionaries, for the Enabling Learning post on building AI language-learning tutors.

A Good LLM Doesn’t Make a Good Teacher

When people learn that our speaking apps run on advanced AI, they ask which model we use. It is the least interesting part of the story. Choosing a capable model is the easy part; any developer can connect to one. What decides whether a learner actually improves is whether the system understands how people learn to speak a second language. That knowledge does not come from the model. It comes from us, from years in bilingual classrooms, instructional design, and second language acquisition research. This post is about how we build AI tutors as educators, not just engineers.