English Language Center Instructors Supercharge Language Study with AI

At The University of Texas at Austin’s English Language Center (ELC), where English is taught as a second or additional language, artificial intelligence is not regarded as a novelty. Instead, it has become infrastructure, intentionally and thoughtfully embedded in lesson planning, student practice and teaching methodologies. 

Sarah Episcopo is among the ELC’s early AI adopters who are leading the charge to incorporate the judicious, ethical and transparent use of AI into classroom pedagogy to boost efficacy for the instructors teaching English and the students learning from them. 

A 15-year veteran instructor, Episcopo teaches academic writing for the ELC’s Academic English Program (AEP), including a summer course directed at students newly enrolled at UT, and a speaking and pronunciation course co-presented with Texas Law for international lawyers-to-be. She also has facilitated AI courses and workshops for summer teacher training groups from Brazil and Taiwan. 

“What our students are doing is amazing: getting graduate degrees in a second language in another country,” said Episcopo. “AI is changing how quickly and efficiently they’re able to do that.” 

Joining Episcopo in AI exploration and implementation are fellow ELC instructors Andrea Holloway, Jill Rolston-Yates and several others. In 2024, Rolston-Yates conducted research into the UT Grammarly AI Pilot program on a research team convened during the University’s Year of AI. Holloway has served on committees assessing alignment of responsible AI implementation with departmental and University guidelines. 

“AI doesn’t replace good teaching or learner effort, but it amplifies both,” said Holloway. “I believe that AI is a great tool for students when paired with guidance from professionals who understand sound pedagogy and learner awareness.” 

ELC instructors do not treat AI as a shortcut that allows students to bypass learning. Instead, they use it as a scaffold — to expand access, personalize instruction and make language practice more frequent, flexible and reflective, without lowering academic standards. 

Access to Practice

Language classrooms are heterogeneous by nature. Students arrive with different linguistic backgrounds, academic goals and strengths, even when placed at the same language proficiency level. AI allows instructors to respond to that variability without diluting rigor by adapting reading materials to different levels, generating contextualized examples of a vocabulary item or explaining grammatical concepts with varying degrees of complexity. 

Another of the most immediate benefits AI offers English learners is on-demand access to practice and feedback beyond classroom time and space. Students can obtain instant responses to questions about vocabulary, grammar, tone or pronunciation, tailored to their proficiency level. 

Speaking aloud is often the most emotionally fraught skill for English learners. Fear of making mistakes, particularly in front of peers, can inhibit experimentation. AI allows students to rehearse and get helpful AI feedback privately in a low-stakes environment that encourages risk taking, an essential component of language development. 

Holloway, who has taught writing and grammar since 2015 in the ELC’s English Language Program (ELP), Academic English Program (AEP) and customized programs, recently created a pronunciation app that now serves as a study helper and asynchronous course for students across multiple time zones. 

“I was teaching a course for employees in Tokyo, Seoul and Shanghai, and the time difference was getting to me,” she recalled. 

The app begins with an AI-generated “podcast” built from Holloway’s lectures. Students listen, review accompanying materials, practice speaking in response to prompts and then receive feedback. AI functions as an adaptive reference tool that responds to follow-up questions in ways static resources cannot. 

More broadly, ELC students can use AI tools as pronunciation models, conversation partners and role-play facilitators: asking for clarification, requesting repetition or focusing on intonation and emphasis, all at their own pace. The result is more speaking practice overall and greater confidence when it comes time for live interaction. 

Discernment as Prime Objective 

A recurring concern about AI in education is the temptation to cede responsibility for original thinking and “let the tool do the work.” Alongside promoting AI’s benefits, ELC instructors address that risk directly. 

Rolston-Yates, who has taught ELP, AEP and Winter International Academy classes for 4 years, created the curriculum for and taught “Holistic Course Design” for one of the center’s specialized teacher training programs.

“I believe that if AI isn’t used responsibly, then it can hinder learning,” she said. “One issue is the ‘cognitive offloading’ that can occur when students depend on AI to do their creative work [such as] writing and other producing that engages critical thinking skills.” 

To counter this, instructors explore the risks and outcomes with students and establish clear boundaries. In writing courses, students typically draft assignments without AI assistance, sometimes through multiple iterations. AI may be used for organization beforehand or error analysis afterward. 

“I allow them to use it at the brainstorming and editing stages of writing, but not at the drafting stage,” said Rolston-Yates. “They use AI to analyze their drafts and search for errors, such as run-on sentences or fragments, so they can edit.” 

Rather than producing machine-polished prose, this approach heightens students’ awareness of their own writing patterns, such as recurring grammatical issues, weak cohesion or tonal inconsistency.  

Rolston-Yates also sets guidelines around accountability and privacy, encouraging students to use UT-endorsed tools such as Copilot and UT Spark. She discusses broader implications, including environmental costs and data privacy, fostering an atmosphere of informed, measured adoption. 

“This is the goal,” said Episcopo. “To help students take pride in how they’re using AI and when they’re using AI, and to help them discern when not to use AI.” 

Teaching Efficiently, Creatively 

AI’s benefits extend beyond student work, of course. For instructors, it has become a powerful planning partner and customizable resource library. AI methodologies are now being shared with future teachers in the ELC’s teacher training programs, for which Episcopo leads a “Technology in the Classroom” course. 

“One thing I tell the teacher trainers is that with AI, anything is possible,” she said. “Our task is to dream of what could work, or what could be great, or what could help our students understand something better, and then ask AI to make it for us. Or if we can’t dream it, ask AI to dream it for us.” 

ELC faculty use AI to generate lesson plans, classroom activities, writing models and diagnostic assessments aligned with specific learning targets. Tasks that once took hours can now be completed more efficiently, freeing instructors to focus on pedagogy, mentoring and course design. 

“It’s essentially a teaching aide — a private secretary that’s taking notes and helping me teach better,” Episcopo said. “It’s helping me to be more efficient and more effective. We’re able to move so much more quickly through material now.”  

Some also use AI to map course calendars, identifying where students typically struggle and when more instructional time is needed. Rolston-Yates views AI as a way to make preparation more innovative; Holloway uses it consistently to generate new materials. 

“You always need to doublecheck, proofread and edit,” said Holloway, “but AI can create some really good skeletons of materials that would normally take hours.” 

AI can also benefit teachers in terms of professional development. In recent years, Holloway and Rolston-Yates have presented in Mexico at the annual conference of College of English Language Teaching Professionals and National Association of Universities and Higher Education Institutions, a gathering of English-language professors primarily from universities in Mexico and Latin America. 

The pair’s 2025 presentation focused on AI micro-learning tools in teacher training and professional development, exploring how micro-lessons powered by AI can provide on-demand training adapted to teachers’ individual needs and schedules to make training more accessible, self-paced and motivating for busy educators.  

Ethics of Use and Disclosure 

AI has also opened space for deeper conversations about ethics and authorship, of course. Episcopo sees these discussions as essential to moving beyond fear-based narratives into proficient, intentional use and integration. 

“I ask students, ‘What do you need to know about AI? How can I help you?’” she said. “Many of them are terrified because the public discourse says, ‘Don’t use it!’ And the problem I have with that public discourse is that it doesn’t qualify what ‘use’ means.” 

Rather than banning AI outright, ELC instructors model transparency and ask students to reflect on when AI use is appropriate and when disclosure is necessary. Discernment and openness become teaching tools. 

“Our role as educators is changing because of AI,” Episcopo said. “Our assumptions of what is humanmade and AI-made ... We’re going to start having more and more doubts. I really feel that’s what our students need from us as teachers: That discernment is what they’re looking for.” 

This emphasis aligns with broader University priorities around responsible innovation. UT’s guidelines for the Responsible Adoption of AI for Teaching and Learning emphasize ethical, intentional and transparent use — tenets that echo those put into daily practice at the English Language Center.  

In fact, Episcopo would prefer that public discourse progress past simple discernment into the details of disclosure.   

“Everything made with AI should be disclosed to the consumer, whoever’s consuming that information,” Episcopo said. “And if we feel shame or embarrassment about disclosing, that’s a good indicator that we shouldn’t be using AI for that product. So, my job is now to model what open disclosure looks like.” 

Augmenting Value, Raising the Bar 

Across vocabulary, reading, speaking, writing, planning and assessment, the pattern is consistent: AI works best when it amplifies good teaching rather than replacing it. It’s a worthy goal to reach for, says Episcopo, and one that should not be limited by fear of the unknown. 

“One of my biggest missions is helping transform the stigma — because the stigma, it’s not working,” said Episcopo. “It’s working against us right now, so we’ve got to talk about it and work with it and figure this out together.” 

Approached correctly, AI can expand access without lowering expectations and personalize learning without fragmenting community. It can illuminate the actions of revision, experimentation and reflection that lie at the heart of language acquisition and learning in general. 

“One of the cool things I think AI is doing for us is, it’s increasing the value of human work and human product,” concluded Episcopo. “My students are not afraid anymore of making errors like they were in the past.” 

With a mindset of frank acceptance, curiosity and intentionality, the instructors at UT’s English Language Center have risen to meet the challenge of understanding what AI can offer and sharing that knowledge with their students and colleagues. It is a gift that does not erode the value of human instruction but raises the bar for it.