Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 222

GENERATIVE AI AS CO-TUTOR: ENHANCING TEACHER ACCOMPANIMENT

STRATEGIES IN EFL ONLINE CLASSROOMS

INTELIGENCIA ARTIFICIAL GENERATIVA COMO CO-TUTOR: MEJORANDO LAS

ESTRATEGIAS DE ACOMPAÑAMIENTO DOCENTE EN AULAS VIRTUALES DE

INGLÉS COMO LENGUA EXTRANJERA

Autores: ¹David Gortaire Díaz, ²Erika Mora Herrera, ³Roddy Real Roby y

Gabriela Almache

Granda.

¹ORCID ID:

https://orcid.org/0000-0001-7364-7305

²ORCID ID: https://orcid.org/0000-0002-8156-0557

²ORCID ID: https://orcid.org/0000-0003-1474-9349

ORCID ID:

https://orcid.org/0000-0003-1858-7121

¹E-mail de contacto: dgortaire@utb.edu.ec

²E-mail de contacto: emorah@utb.edu.ec

³E-mail de contacto: rreal@utb.edu.ec

E-mail de contacto:

galmache@utb.edu.ec

Afiliación:

1*2*3*4*

Universidad Técnica de Babahoyo, (Ecuador).

Artículo recibido: 15 de Febrero del 2026

Artículo revisado: 18 de Febrerodel 2026

Artículo aprobado: 23 de Febrero del 2026

¹Ingeniero en Negocios Internacionales, Escuela Superior Politécnica del Litoral, (Ecuador), 12 años de experiencia laboral. Master en

Desarrollo Rural de la Escuela Superior Politécnica del Litoral, (Ecuador), y Master en Pedagogía del Inglés como Lengua Extranjera de

la Universidad Bolivariana del Ecuador, (Ecuador).

²Licenciada en Lengua y Lingüística Inglesa, titulada en la Universidad de Filosofía Ciencia y Letras de la Educación-Escuela de Lengua,

(Ecuador). Maestría en Pedagogía del Idioma Inglés como Lengua Extranjera de la Universidad Bolivariana del Ecuador, (Ecuador).

³Ingeniero en Negocios Internacionales, titulado por la Escuela Superior Politécnica del Litoral (ESPOL), (Ecuador). Máster Universitario

en Dirección de Empresas con mención en Negocios Internacionales por la Universidad de Palermo, (Argentina). Además, posee una

Maestría en Pedagogía del Idioma Inglés como Lengua Extranjera de la Universidad Bolivariana del Ecuador, (Ecuador).

Ingeniera en Ciencias Empresariales, mención en Dirección y Planificación Comercial, Universidad de Especialidades Espíritu Santo,

(Ecuador). Magíster en Pedagogía de los Idiomas Nacionales y Extranjeros, mención en Enseñanza de Inglés, Universidad Casa Grande,

(Ecuador). Magíster en Educación con mención en Innovaciones Pedagógicas, Universidad Casa Grande, (Ecuador). Doctorante en

Educación e Innovación, Universidad de Investigación e Innovación de México, (México).

Resumen

Este estudio examinó la implementación y

efectividad de la inteligencia artificial (IA)

generativa como co-tutor para mejorar las

estrategias de acompañamiento docente en

aulas virtuales de inglés como lengua

extranjera (EFL) en una universidad pública de

Ecuador. La investigación indagó las

percepciones docentes, prácticas de

implementación, desafíos encontrados y

efectos observados en el compromiso

estudiantil tras una intervención colaborativa

de desarrollo profesional que integró IA en

marcos pedagógicos establecidos. Se empleó

un diseño de estudio de caso cualitativo con 26

docentes de EFL del Centro de Idiomas de la

Universidad Técnica de Babahoyo. La

intervención de tres meses (octubre-diciembre

2025) consistió en sesiones de capacitación,

talleres colaborativos y desarrollo participativo

de estrategias. La recolección de datos incluyó

protocolos de reflexión estructurada,

cuestionarios exhaustivos con ítems de escala

Likert y preguntas abiertas, y artefactos

generados por los docentes. El análisis

combinó análisis temático de datos cualitativos

siguiendo el marco de Braun y Clarke (2006)

con análisis estadístico descriptivo de medidas

cuantitativas de percepción. Los resultados

revelaron percepciones docentes

predominantemente positivas sobre la IA como

co-tutor, con 96.2% concordando que la IA

complementa en lugar de reemplazar la

enseñanza humana. Los docentes reportaron

mejoras en la confianza estudiantil (65.4%),

participación en actividades asincrónicas

(73.1%) y calidad de interacciones (73.1%). La

retroalimentación de escritura (92.3%),

explicación gramatical (88.5%) y apoyo de

vocabulario (84.6%). A pesar de los desafíos,

80.8% de los participantes tenían intención de

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 223

continuar la implementación de co-tutoría con

IA, demostrando fuerte compromiso con la

integración sostenida. Esta investigación

contribuye evidencia empírica sobre

implementación de co-tutoría con IA en un

contexto latinoamericano de universidad

pública poco estudiado, expandiendo la

diversidad geográfica más allá de entornos de

investigación predominantemente anglófonos.

El estudio avanza marcos teóricos al

reconceptualizar el acompañamiento docente

en educación en línea como sistema distribuido

que combina experticia humana con

capacidades computacionales.

Palabras clave: Inteligencia artificial

generativa, Co-tutoría, Acompañamiento

docente, Educación EFL en línea.

Abstract

This study examined the implementation and

effectiveness of generative artificial

intelligence (AI) as co-tutor to enhance teacher

accompaniment strategies in English as a

Foreign Language (EFL) online classrooms at

a public university in Ecuador. The research

investigated teacher perceptions,

implementation practices, challenges

encountered, and observed effects on student

engagement following a collaborative

professional development intervention

integrating AI into established pedagogical

frameworks. A qualitative case study design

was employed with 26 EFL teachers from the

Language Center at Universidad Técnica de

Babahoyo. The three-month intervention

(October-December 2025) consisted of training

sessions, collaborative workshops, and

participatory strategy development. Data

collection included structured reflection

prompts, comprehensive questionnaires with

Likert-scale and open-ended items, and

teacher-generated artifacts. Analysis combined

thematic analysis of qualitative data following

Braun and Clarke's (2006) framework with

descriptive statistical analysis of quantitative

perception measures. Results revealed

predominantly positive teacher perceptions of

AI as co-tutor, with 96.2% agreeing that AI

complements rather than replaces human

teaching. Teachers reported improvements in

student confidence (65.4%), participation in

asynchronous activities (73.1%), and quality of

interactions (73.1%). Writing feedback

(92.3%), grammar explanation (88.5%), and

vocabulary support (84.6%). Despite

challenges, 80.8% of participants intended to

continue AI co-tutoring implementation,

demonstrating strong commitment to sustained

integration. This research contributes empirical

evidence on AI co-tutoring implementation in

an understudied Latin American public

university context, expanding geographical

diversity beyond predominantly Anglophone

research settings. The study advances

theoretical frameworks by reconceptualizing

teacher accompaniment in online education as

distributed system combining human expertise

with computational capabilities.

Keywords: Generative artificial intelligence,

Co-tutoring, Teacher accompaniment, EFL

online education.

Sumário

Este estudo examinou a implementação e a

eficácia da inteligência artificial (IA)

generativa como cotutora para aprimorar as

estratégias de apoio ao professor em salas de

aula virtuais de inglês como língua estrangeira

(EFL) em uma universidade pública no

Equador. A pesquisa explorou as percepções

dos professores, as práticas de implementação,

os desafios encontrados e os efeitos observados

no engajamento dos alunos após uma

intervenção colaborativa de desenvolvimento

profissional que integrou a IA a estruturas

pedagógicas estabelecidas. Utilizou-se um

estudo de caso qualitativo com 26 professores

de inglês como língua estrangeira do Centro de

Línguas da Universidade Técnica de

Babahoyo. A intervenção, com duração de três

meses (outubro a dezembro de 2025), consistiu

em sessões de treinamento, oficinas

colaborativas e desenvolvimento participativo

de estratégias. A coleta de dados incluiu

protocolos de reflexão estruturados,

questionários abrangentes com itens em escala

Likert e questões abertas, além de artefatos

gerados pelos professores. A análise combinou

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 224

a análise temática dos dados qualitativos,

seguindo a estrutura de Braun e Clarke (2006),

com a análise estatística descritiva das medidas

quantitativas de percepção. Os resultados

revelaram percepções predominantemente

positivas do corpo docente em relação à IA

como cotutora, com 96,2% concordando que a

IA complementa, e não substitui, o ensino

humano. Os docentes relataram melhorias na

confiança dos alunos (65,4%), na participação

em atividades assíncronas (73,1%) e na

qualidade das interações (73,1%). Também

foram relatados feedback sobre a escrita

(92,3%), explicações gramaticais (88,5%) e

suporte ao vocabulário (84,6%). Apesar dos

desafios, 80,8% dos participantes pretendem

continuar implementando a cotutoria com IA,

demonstrando um forte compromisso com a

integração sustentada. Esta pesquisa contribui

com evidências empíricas sobre a

implementação da cotutoria com IA em um

contexto universitário público latino-

americano relativamente inexplorado,

expandindo a diversidade geográfica para além

dos ambientes de pesquisa predominantemente

anglófonos. Este estudo avança os referenciais

teóricos ao reconceitualizar o apoio ao

professor na educação online como um sistema

distribuído que combina expertise humana com

capacidades computacionais.

Palavras-chave: Inteligência artificial

generativa, Cotutoria, Apoio ao professor,

Educação de inglês como língua estrangeira

online.

Introduction

The advancement of artificial intelligence

technologies has developed unprecedented

transformations across educational

environment, with generative AI emerging as a

particularly disruptive force in language

learning contexts (Celik, 2023). Within the

domain of English as a Foreign Language (EFL)

instruction, the integration of AI-powered tools

has evolved from peripheral supplementary

resources to central pedagogical instruments

capable of fundamentally reshaping

instructional methodologies and learner

engagement patterns (Lara et al., 2023; Sun et

al., 2021). This paradigm shift has intensified

following the proliferation of online learning

environments, where the absence of physical

classroom presence and limited real-time

instructor availability have created pronounced

challenges in maintaining consistent learner

support and personalized guidance (Pikhart,

2020). The COVID-19 pandemic accelerated

this transition, exposing critical gaps in

traditional teacher accompaniment strategies

while simultaneously revealing opportunities

for technological intervention to enhance

pedagogical effectiveness in virtual settings

(Popenici & Kerr, 2017).

Generative AI systems, particularly large

language models such as ChatGPT, Claude, and

specialized educational platforms, possess

unique capabilities that align with the

multifaceted demands of EFL instruction,

including instantaneous feedback provision,

adaptive scaffolding, personalized content

generation, and sustained learner engagement

beyond synchronous class sessions (Nazari et

al., 2021). These technological affordances

present compelling possibilities for

reconceptualizing the instructor's role from sole

knowledge provider to orchestrator of AI-

enhanced learning ecosystems, wherein

generative AI functions as a co-tutor that

extends, rather than replaces, human

pedagogical expertise (Keezhatta, 2019).

However, the theoretical frameworks

underpinning effective human-AI collaboration

in educational contexts remain nascent, and

empirical evidence regarding optimal

integration strategies, particularly within EFL

online environments, continues to develop

(Fahimirad & Kotamjani, 2018).

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 225

The concept of teacher accompaniment, deeply

rooted in sociocultural learning theories and

constructivist pedagogical approaches,

emphasizes sustained guidance, emotional

support, and adaptive responsiveness to

individual learner trajectories throughout the

educational process (Zhang & Lin, 2018). In

traditional face-to-face EFL contexts, this

accompaniment manifests through immediate

error correction, contextual explanations,

motivational encouragement, and progressive

complexity adjustment based on observable

learner performance (Fernandez et al., 2013).

Online learning modalities, however, introduce

temporal and spatial discontinuities that

complicate these accompaniment processes,

frequently resulting in diminished learner

autonomy, increased anxiety, and reduced

persistence when confronted with linguistic

challenges during asynchronous study periods

(Ercan, 2018). Generative AI technologies offer

potential solutions to these limitations by

providing continuous availability, immediate

responsiveness, and individualized support

mechanisms that can bridge gaps between

synchronous instructional sessions (Zhai &

Wibowo, 2023).

Despite the promising potential of generative

AI as pedagogical support, critical concerns

persist regarding implementation effectiveness,

pedagogical alignment, learner autonomy

preservation, ethical considerations, and the

maintenance of essential human elements in

language acquisition processes (Zhai &

Wibowo, 2023). Questions emerge regarding

optimal task distribution between human

instructors and AI systems, the development of

learner competencies for effective AI

interaction, potential overdependence on

automated support systems, accuracy and

appropriateness of AI-generated linguistic

feedback, and the preservation of sociocultural

dimensions essential to communicative

language development (Arini et al., 2022).

Furthermore, the integration of generative AI

within existing pedagogical frameworks

necessitates careful consideration of how such

tools complement established teaching

methodologies, curriculum objectives, and

assessment practices without undermining core

instructional values or compromising learning

outcomes (Salas & Yang, 2022).

This investigation examines the implementation

and effectiveness of generative AI as a co-

tutoring mechanism within EFL online

classrooms, specifically focusing on how these

technologies can enhance traditional teacher

accompaniment strategies while maintaining

pedagogical integrity and fostering authentic

language acquisition. Through systematic

analysis of integration models, learner

engagement patterns, instructional design

considerations, and outcome measures, this

study seeks to establish evidence-based

frameworks for optimizing human-AI

collaboration in virtual EFL learning

environments. The research addresses critical

gaps in current literature by examining practical

implementation challenges, identifying

effective practices for AI-mediated support, and

proposing theoretical models that position

generative AI as complementary rather than

substitutive educational technology within the

broader ecosystem of online language

instruction.

Materials and Methods

This study employed a qualitative case study

design to examine the implementation and

effectiveness of generative AI as co-tutor in

enhancing teacher accompaniment strategies

within EFL online classrooms. The study was

conducted at the Centro de Idiomas (Language

Center) of Universidad Técnica de Babahoyo, a

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 226

public university located in the Los Ríos

province of Ecuador. The Language Center

serves students from across the university's

academic programs, providing required English

instruction as part of general education

requirements alongside optional advanced

courses and certification preparation programs.

Classes are conducted primarily through online

modalities, a configuration accelerated by

pandemic-related institutional transformations

and maintained due to geographical

accessibility challenges and resource

optimization considerations. The research

participants comprised 26 EFL teachers

employed at the Language Center during the

October-December 2025 period when the study

was conducted. This sample represented the

complete population of active instructors at the

center during the study period, constituting a

census sampling approach. Participant

demographics reflected diversity in teaching

experience, ranging from early-career

instructors with fewer than three years of

experience to veteran educators with over

fifteen years in EFL instruction. Participation in

the study was voluntary, though all Language

Center instructors were invited and encouraged

to engage given the institutional priority placed

on pedagogical innovation and professional

development.

The intervention consisted of a three-phase

professional development program designed

collaboratively with Language Center

leadership to integrate generative AI co-tutoring

strategies into existing EFL online instruction.

The intervention extended across the October-

December 2025 period, encompassing initial

training sessions, collaborative workshop

activities, and ongoing implementation support

as teachers experimented with AI integration in

their courses. Phase One, conducted during

October 2025, consisted of foundational

training sessions introducing participants to

generative AI technologies, their capabilities

and limitations, and potential applications for

language teaching. Training was delivered

through synchronous online workshops

utilizing modeling, demonstration, and guided

practice.

Phase Two, spanning November 2025, involved

collaborative workshops where participants

worked in small groups to design specific AI

co-tutoring strategies and tools aligned with

their instructional contexts and student needs.

Workshop activities included identification of

accompaniment challenges in current online

teaching practices, brainstorming of AI-

supported solutions, development of student-

facing AI prompts and guidance materials, and

peer review of proposed strategies. Phase

Three, conducted throughout December 2025,

involved implementation of designed strategies

in participants' actual courses with ongoing

support and reflective practice activities.

Teachers were encouraged to experiment with

various AI co-tutoring configurations,

document their experiences, and share insights

through informal peer discussions and

structured reflection protocols.

Data collection employed multiple methods to

capture diverse dimensions of teacher

experiences, perceptions, and practices related

to AI co-tutoring implementation. Data

collection occurred concurrently with

intervention implementation, allowing capture

of real-time experiences and evolving

perceptions as participants progressed through

training, strategy development, and classroom

implementation phases. Primary data sources

included structured reflection prompts

completed by participants at key junctures

throughout the intervention period.

Additionally, participants completed a

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 227

comprehensive end-of-intervention

questionnaire examining multiple dimensions

of their AI co-tutoring experiences and

perceptions. The questionnaire incorporated

both closed-ended items utilizing Likert-scale

response formats and open-ended questions

inviting detailed explanations and examples.

Closed-ended items addressed constructs

including perceived usefulness of AI as co-

tutor, ease of integration into existing

pedagogical practices, observed changes in

student engagement, confidence in using AI

tools, and intentions for continued use beyond

the study period.

Qualitative data from reflection prompts, open-

ended questionnaire responses, and artifacts

underwent thematic analysis. Initial

familiarization involved reading all textual data

multiple times to develop comprehensive

understanding and noting preliminary

observations. Quantitative data from closed-

ended questionnaire items were analyzed using

descriptive statistics including frequency

distributions, measures of central tendency, and

measures of variability. Descriptive statistical

analysis was conducted using SPSS software,

with results presented through tables and graphs

facilitating interpretation and comparison

across different questionnaire items. While the

relatively small sample size precluded robust

inferential statistical analyses, descriptive

statistics offered valuable information about the

distribution and central tendencies of teacher

perceptions within this specific case context.

Results and Discussion

The implementation of generative AI as co-

tutor within the EFL online classrooms at

Universidad Técnica de Babahoyo's Language

Center yielded multifaceted findings regarding

teacher perceptions, implementation practices,

and observed effects on student engagement.

Data collected across the three-month

intervention period (October-December 2025)

through structured reflections, questionnaires,

and collaborative workshop activities revealed

predominantly positive teacher attitudes toward

AI co-tutoring alongside recognition of

significant challenges requiring careful

navigation.

Table 1. Demographic and Professional

Characteristics of Study Participants (N = 26)

Characteristic

Category

Never

n (%)

Rarely

n (%)

Sometimes

n (%)

Often

n (%)

Very

Often

n (%)

Technical

Challenges

Students

lack access

to AI tools

(30.8)

(38.5)

6 (23.1)

2 (7.7)

0 (0.0)

2.08

0.93

AI platforms

experiencing

downtime or

errors

(11.5)

(46.2)

9 (34.6)

2 (7.7)

0 (0.0)

2.38

0.80

Difficulty

integrating

AI into LMS

(19.2)

(34.6)

8 (30.8)

(15.4)

0 (0.0)

2.42

0.99

Pedagogical

Challenges

AI generates

inaccurate or

inappropriate

responses

1 (3.8)

(26.9)

13 (50.0)

(19.2)

0 (0.0)

2.85

0.78

Difficulty

designing

effective

prompts

2 (7.7)

(42.3)

9 (34.6)

(15.4)

0 (0.0)

2.58

0.84

Students

over-rely on

AI without

learning

0 (0.0)

(19.2)

10 (38.5)

(34.6)

2 (7.7)

3.31

0.88

Students use

AI to

complete

work

without

effort

0 (0.0)

(15.4)

9 (34.6)

(38.5)

(11.5)

3.46

0.90

Difficulty

monitoring

student AI

use

1 (3.8)

(23.1)

8 (30.8)

(34.6)

2 (7.7)

3.19

1.02

Training

and

Support

Challenges

Insufficient

time to learn

AI tools

(11.5)

(30.8)

10 (38.5)

(19.2)

0 (0.0)

2.65

0.94

Students

need more

guidance on

AI use

0 (0.0)

(11.5)

8 (30.8)

(46.2)

(11.5)

3.58

0.86

Lack of

institutional

guidelines

on AI

2 (7.7)

(19.2)

7 (26.9)

(30.8)

(15.4)

3.27

1.18

Ethical and

Policy

Challenges

Concerns

about

academic

integrity

1 (3.8)

(23.1)

9 (34.6)

(30.8)

2 (7.7)

3.15

0.97

Privacy and

data security

concerns

(15.4)

(38.5)

8 (30.8)

(11.5)

1 (3.8)

2.50

1.00

Equity issues

due to

unequal

access

(23.1)

(34.6)

7 (26.9)

(11.5)

1 (3.8)

2.38

1.06

Source: Own elaboration

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 230

Table 5 presents the frequency with which

participants encountered various challenges

across technical, pedagogical, training, and

ethical dimensions, revealing that pedagogical

concerns predominated over technical

obstacles. Among pedagogical challenges, the

most frequently encountered issue involved

students using AI to complete work without

genuine effort (M = 3.46, SD = 0.90), with

50.0% of teachers reporting this occurred often

or very oftenClosely related was concern about

student over-reliance on AI without actual

learning (M = 3.31, SD = 0.88), reported as

occurring sometimes, often, or very often by

80.8% of participants. These patterns

underscore the critical importance of explicit

instruction about strategic AI use and

assessment design that requires demonstration

of understanding rather than mere production of

correct answers. Teachers also frequently

encountered difficulty monitoring student AI

use (M = 3.19, SD = 1.02) and navigating

students' need for more guidance on appropriate

AI use (M = 3.58, SD = 0.86), the latter

occurring often or very often for 57.7% of

participants. Finally, Table 6 presents prompts

developed by teacher that could be useful for

other academics,

Table 6. Collaboratively Designed AI Prompts for EFL Student Use Across Learning Activities (N =

26)

Learning Activity

Prompt Category

Designed Prompt Template

Pedagogical Rationale

Developed by

n Teachers

Writing Practice

Essay Feedback

"I am an [proficiency level] English learner writing an essay about [topic]. Please read my draft below and provide feedback by: 1)

Identifying three strengths in my writing, 2) Explaining two areas where I can improve with specific examples from my text, 3)

Asking me questions to help me think deeper about my ideas. Do NOT rewrite my essay. Help me understand how to improve it

myself. Here is my draft: [student paste draft]"

Encourages reflective revision rather

than passive acceptance of corrections;

maintains student ownership of writing

process

Writing Practice

Grammar Self-

Correction

"I am practicing English grammar at [level]. I wrote this sentence: '[student sentence]'. I think there might be a grammar error related

to [student's hypothesis, e.g., verb tense]. Can you: 1) Tell me if my hypothesis is correct, 2) Explain the grammar rule in simple

terms, 3) Show me how to correct it, 4) Give me two similar practice sentences to try?"

Promotes metalinguistic awareness by

requiring students to hypothesize about

errors before receiving answers

Speaking Practice

Conversation

Partner

"Act as a conversation partner for an English learner at [level]. Start a conversation about [topic of interest]. Keep your responses to

2-3 sentences. Ask me follow-up questions based on what I say. If I make grammar or vocabulary mistakes, gently correct me by

restating what I said correctly, then continue the conversation. Adjust your vocabulary to match my level."

Provides low-anxiety speaking practice

with natural conversation flow and

implicit error correction

Speaking Practice

Pronunciation

Guidance

"I am learning English pronunciation. I want to practice saying: '[target word/phrase]'. Can you: 1) Break down the word into

syllables and show me the stress pattern, 2) Describe how to position my mouth and tongue for difficult sounds, 3) Give me similar

words to practice with the same sound patterns, 4) Suggest tongue twisters or sentences for practice?"

Offers detailed phonetic guidance

unavailable in traditional text-based

materials

Vocabulary

Development

Contextual Usage

"I am learning the English word '[target word]' at [level]. Please help me understand it by: 1) Defining it in simple English, 2)

Showing it in three different example sentences that demonstrate different contexts, 3) Teaching me 2-3 common collocations with

this word, 4) Telling me if this word is formal, informal, or neutral, 5) Giving me a short story (4-5 sentences) that uses this word

naturally."

Moves beyond dictionary definitions to

develop rich, contextualized word

knowledge

Vocabulary

Development

Synonym

Differentiation

"I know the English word '[word 1]' but I learned a similar word '[word 2]'. I am confused about when to use each one. Can you: 1)

Explain the difference in meaning between them, 2) Tell me which situations are appropriate for each word, 3) Give me example

sentences showing the difference, 4) Create a practice exercise where I choose between them?"

Addresses common confusion between

near-synonyms requiring nuanced

understanding

Reading

Comprehension

Text Scaffolding

"I am reading this English text at [level]: '[paste text]'. Before I answer my teacher's questions, please help me understand it better by:

1) Summarizing the main idea in one simple sentence, 2) Identifying 3-5 key vocabulary words I should understand, 3) Explaining

any cultural references or idioms, 4) Asking me 2-3 comprehension questions to check my understanding. Do NOT answer my

teacher's assignment questions for me."

Provides comprehension support while

maintaining academic integrity by

avoiding direct assignment completion

Reading

Comprehension

Vocabulary in

Context

"I am reading an English text and I don't understand this sentence: '[paste sentence]'. The word '[target word]' is confusing to me. Can

you: 1) Explain what this word means in THIS specific sentence, 2) Show me if this word has other meanings in different contexts, 3)

Rewrite the sentence in simpler English keeping the same meaning, 4) Help me understand why the author chose this particular

word?"

Develops context-based meaning

inference skills rather than isolated

vocabulary memorization

Grammar

Practice

Rule Explanation

"I am learning about [grammar structure, e.g., present perfect tense] in English at [level]. Can you: 1) Explain this grammar in simple

terms with the basic rule, 2) Tell me when native speakers use this grammar, 3) Show me 5 example sentences, 4) Explain common

mistakes learners make with this grammar, 5) Give me a short practice exercise (5 questions) and then check my answers with

explanations?"

Provides comprehensive grammar

instruction with practice opportunities

and immediate feedback

Grammar

Practice

Error Pattern

Analysis

"I keep making the same grammar mistake in my writing. Here are three sentences where my teacher marked errors: [paste

sentences]. Can you: 1) Identify what grammar pattern I am struggling with, 2) Explain the correct rule, 3) Show me the corrected

sentences, 4) Create personalized practice exercises targeting my specific problem, 5) Give me strategies to remember this rule?"

Addresses individual error patterns with

targeted intervention

Cultural

Competence

Pragmatic

Appropriateness

"I want to say '[student's intended message]' in English to [describe situation and relationship, e.g., 'my professor in an email']. Is my

way of saying this appropriate for this situation? Can you: 1) Tell me if my phrasing is too formal, too informal, or appropriate, 2)

Explain why, 3) Suggest alternative ways to express the same idea with different levels of formality, 4) Teach me cultural norms for

this type of communication in English-speaking countries?"

Develops sociolinguistic competence

and pragmatic awareness

Cultural

Competence

Idiomatic

Expression

"I heard this English expression: '[idiom/colloquialism]' and I don't understand it. Can you: 1) Explain its literal meaning versus its

actual meaning, 2) Tell me in what situations people use this expression, 3) Give me examples in sentences, 4) Teach me if this is

formal or casual language, 5) Share similar expressions with the same meaning?"

Bridges cultural gaps in understanding

non-literal language

Exam

Preparation

Practice Test

Creation

"I am preparing for [specific exam, e.g., TOEFL writing section]. I need to practice [skill area]. Can you: 1) Create a practice question

similar to the real exam, 2) Give me 20 minutes to complete it, 3) After I submit my answer, evaluate it using the official exam rubric,

4) Provide specific feedback on strengths and areas for improvement, 5) Suggest strategies for better performance?"

Simulates authentic exam conditions

with constructive feedback

Exam

Preparation

Strategy

Development

"I am taking [exam name] and I struggle with [specific section/skill]. Can you: 1) Explain common question types in this section, 2)

Teach me time management strategies, 3) Share tips for avoiding common mistakes, 4) Give me a step-by-step approach for

answering these questions, 5) Create mini-practice exercises for each strategy you teach me?"

Develops test-taking strategies

alongside language skills

Listening

Practice

Transcript

Analysis

"[After student listens to audio] I listened to an English audio about [topic]. Here is what I understood: [student summary]. Can you

help me check my comprehension by: 1) Confirming what I understood correctly, 2) Pointing out important information I missed, 3)

Explaining vocabulary or phrases I might not have caught, 4) Asking follow-up questions to deepen my understanding? Do NOT give

me the transcript unless I specifically ask for it."

Encourages active listening and

comprehension checking without

immediately providing answers

Self-Assessment

Learning Progress

Reflection

"I am learning English at [level]. This week I practiced [activities completed]. Can you help me reflect on my progress by: 1) Asking

me questions about what I found easy and difficult, 2) Helping me identify patterns in my mistakes, 3) Suggesting what I should focus

on next week, 4) Creating a personalized mini study plan based on my specific needs?"

Promotes metacognitive awareness and

learner autonomy

Translation

Analysis

Contrastive

Analysis

"In my native language [specify language], we say '[phrase in L1]' to express [meaning/function]. In English, I translated it as

'[student's English translation]'. Can you: 1) Tell me if my English translation captures the same meaning and tone, 2) Explain any

differences between how my language and English express this idea, 3) Suggest more natural English ways to say this, 4) Teach me

about any cultural differences in how this concept is expressed?"

Uses L1 as resource for contrastive

analysis while avoiding over-reliance

on direct translation

Motivation

Support

Learning Strategy

Coaching

"I am feeling [frustrated/overwhelmed/discouraged] about learning English because [specific challenge]. Can you: 1) Help me

understand that this challenge is normal for language learners, 2) Suggest 3-4 specific strategies to overcome this challenge, 3) Share

encouraging facts about language learning, 4) Help me set a small, achievable goal for this week, 5) Remind me of progress I've

likely already made?"

Provides emotional support and

strategic guidance during challenging

learning periods

Source: Own elaboration

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 231

The findings from this case study provide

empirical evidence that generative AI can

function effectively as co-tutor within EFL

online classrooms, enhancing teacher

accompaniment strategies while simultaneously

revealing significant pedagogical, technical,

and ethical challenges requiring careful

research (Hwang et al., 2020; Salas & Yang,

2022). The predominantly positive teacher

perceptions documented in this research,

particularly regarding AI's capacity to extend

temporal boundaries of instructional support

and provide immediate feedback, align with

theoretical predictions from distributed

cognition frameworks suggesting that effective

learning environments can emerge from

strategic coordination between human expertise

and computational capabilities (Sumakul et al.,

2022).

However, the nuanced pattern of

findings;where teachers simultaneously

endorsed AI's value while expressing concerns

about academic integrity, accuracy limitations,

and potential erosion of human

connection,underscores that successful AI

integration depends not merely on technological

capability but on thoughtful pedagogical

orchestration that preserves essential human

elements while leveraging computational

affordances (Braiki et al., 2020; Jiang, 2022).

These results contribute to emerging literature

on educational AI by documenting

implementation processes and perception

patterns within an understudied Latin American

public university context, thereby expanding

geographical and institutional diversity of

empirical evidence beyond predominantly

Anglophone and East Asian settings that have

dominated existing research (Chen et al., 2020).

The strong agreement among participants that

AI complements rather than replaces human

teaching (96.2% agreement) represents a

particularly significant finding, suggesting that

the intervention's deliberate framing of AI as

collaborative tool rather than instructor

substitute successfully mitigated potential

resistance rooted in professional identity

threats. This outcome aligns with (Wang, 2022)

argument that educational technologies should

be designed for empowerment rather than

replacement, supporting human capabilities

instead of substituting them. The collaborative,

participatory design approach employed in this

study; where teachers actively shaped AI

integration strategies rather than receiving top-

down mandates, likely contributed to this

positive framing by positioning educators as

professional decision-makers exercising agency

over technology adoption (Wang, 2019).

The diversity of AI co-tutoring strategies

implemented by participants, ranging from

highly adopted applications like writing

feedback (92.3%) and grammar explanation

(88.5%) to less common uses such as cultural

pragmatics guidance (53.8%) and motivational

support (46.2%), reveals how teachers

exercised professional judgment in selecting

applications aligned with their pedagogical

priorities and students' needs (Lara et al., 2023).

The pattern wherein frequently implemented

strategies also received highest effectiveness

ratings suggests that teachers gravitated toward

applications where AI demonstrated clear

capabilities while showing appropriate caution

about domains requiring nuanced cultural

understanding or emotional sensitivity (Sun et

al., 2021). The finding that pedagogical

challenges predominated over technical

obstacles represents an important contribution

to educational technology literature, which

sometimes overemphasizes infrastructure and

access issues while underattending to more

fundamental questions about appropriate use

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 232

and pedagogical integration (Lu, 2019).

However, this finding should not be interpreted

as inherent limitation of AI technology but

rather as indicator that implementation requires

explicit instruction about strategic use,

assessment design preventing gaming, and

cultivation of learning dispositions valuing

understanding over mere task completion

(Feuerriegel et al., 2022). The collaboratively

designed prompts documented in Table 6

represent a significant practical contribution,

demonstrating how pedagogical principles can

be embedded within AI interaction structures to

scaffold appropriate use and maintain learning

objectives. The emphasis in these prompts on

encouraging student thinking rather than

providing direct answers, requiring students to

articulate hypotheses before receiving

corrections, and explicitly prohibiting

completion of teacher-assigned work reflects

sophisticated understanding that AI's

educational value depends critically on how

interactions are structured (Wang, 2022;

Zawacki et al., 2019b).

The integration of AI co-tutoring within the

theoretical framework of teacher

accompaniment in online education represents

an important conceptual contribution of this

research. Traditional conceptualizations of

teaching presence (Kumar et al., 2023) and

teacher accompaniment (Gutierrez et al., 2022)

focused exclusively on human instructor

actions, treating technology as passive medium

rather than active participant in pedagogical

relationships. This study's findings suggest

value in expanding these frameworks to

acknowledge AI as potential co-participant in

distributed accompaniment systems, where

continuous support emerges not from heroic

individual teacher effort to be always available

but from strategic orchestration of human

expertise during synchronous interactions and

AI support during asynchronous periods

(Canhoto & Clear, 2020). This

reconceptualization aligns with ecological

perspectives on educational technology viewing

learning environments as complex systems

where multiple actors; human and

computational,interact to create emergent

educational experiences (Zawacki et al.,

2019a). However, this expanded framework

must preserve recognition that certain

dimensions of accompaniment require uniquely

human capabilities. The qualitative data

revealing teacher concerns about loss of human

connection, combined with moderate

effectiveness ratings for AI-provided

motivational and emotional support, suggest

that socio-emotional accompaniment remains

primarily human domain (Haryanto & Ali,

2019). Future research should examine how to

optimize complementarity between AI's

strengths in immediate, consistent, scalable

support and human teachers' capacities for

empathy, contextual understanding,

relationship building, and nuanced pedagogical

judgment, potentially developing design

principles for hybrid accompaniment systems

that strategically deploy each actor's distinctive

capabilities.

Conclusiones

This case study investigation of generative AI

as co-tutor in EFL online classrooms at

Universidad Técnica de Babahoyo provides

compelling evidence that AI technologies can

meaningfully enhance teacher accompaniment

strategies when implemented through

collaborative, pedagogically grounded

approaches that position technology as

complement rather than substitute for human

expertise. The predominantly positive teacher

perceptions documented across multiple

dimensions—particularly regarding AI's

capacity to extend temporal boundaries of

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 233

support, provide immediate feedback, and

enhance learning personalization—demonstrate

that generative AI possesses genuine

affordances for addressing persistent challenges

in online language education, including limited

opportunities for individualized attention,

delayed feedback, and insufficient practice

beyond synchronous class sessions (Borup et

al., 2020; Hrastinski, 2019). However, the

nuanced pattern of findings, wherein

enthusiasm coexisted with substantial concerns

about academic integrity, accuracy limitations,

student over-reliance, and potential erosion of

essential human connection, underscores that

realizing AI's educational potential requires far

more than technological deployment; it

demands thoughtful pedagogical orchestration,

explicit student preparation for critical and

strategic use, ongoing teacher professional

development, and institutional policies

providing clear guidance while preserving

educator agency (Haryanto & Ali, 2019;

Zawacki et al., 2019a).

The collaborative design approach employed in

this study, where teachers actively shaped

integration strategies through participatory

workshops rather than receiving prescriptive

implementation mandates, proved instrumental

in fostering positive attitudes and sustainable

adoption intentions. The diversity of AI co-

tutoring applications developed by

participants—from writing feedback and

grammar explanation to conversational practice

and cultural pragmatics guidance—illustrates

how professional educators, when provided

appropriate support and autonomy, exercise

sophisticated judgment about which

technological affordances align with their

pedagogical values and students' needs. The

collection of carefully designed prompts created

through this collaborative process represents a

significant practical contribution,

demonstrating how pedagogical principles can

be embedded within AI interaction structures to

scaffold appropriate use, maintain cognitive

engagement, and prevent superficial task

completion without genuine learning. These

prompts exemplify what effective human-AI

collaboration in education might entail:

strategic division of labor where computational

capabilities handle immediate, scalable support

functions while human expertise focuses on

higher-order pedagogical activities requiring

contextual understanding, emotional

sensitivity, and nuanced professional judgment

(Luckin et al., 2016; Roll & Wylie, 2016).

The challenges documented in this research;

particularly the prevalence of concerns about

students using AI to complete work without

effort, difficulties monitoring AI use, and AI's

occasional generation of inaccurate or culturally

inappropriate responses, highlight that current

implementations remain imperfect and require

ongoing refinement. These findings should not

be interpreted as reasons to reject AI integration

but rather as indicators of where additional

effort is needed: developing assessment

approaches that require demonstration of

understanding rather than mere production of

correct answers, cultivating critical AI literacy

among both students and teachers, establishing

institutional guidelines providing clear

direction on appropriate use, and ensuring

equitable access so that technological

advantages do not accrue disproportionately to

already-privileged students (Hwang &

Nurtantyana, 2022). The study's grounding in

an Ecuadorian public university context

provides important empirical evidence that

meaningful AI integration is achievable even in

resource-constrained settings characteristic of

many developing country institutions, provided

that implementation is context-responsive and

realistic about limitations rather than attempting

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 234

to replicate models developed in resource-rich

environments.

This finding has significant implications for

educational equity at both institutional and

global levels, suggesting that thoughtfully

designed AI co-tutoring could help address

quality gaps in contexts where teacher-student

ratios are high and resources for individualized

support are limited, though careful attention to

access barriers and digital divides remains

essential (Selwyn et al., 2020). The strong

teacher intentions for continued use (80.8%

planning to continue implementation) and

willingness to recommend AI co-tutoring to

colleagues (88.5%) provide encouraging

evidence about sustainability prospects, though

longitudinal research will be necessary to

determine whether initial enthusiasm translates

into durable practice transformation. As

generative AI technologies continue their rapid

evolution, the educational community faces

critical decisions about how to harness their

capabilities while preserving essential human

elements and addressing legitimate concerns

about equity, integrity, and learning quality.

This research suggests that productive pathways

forward involve neither uncritical technology

adoption nor reflexive rejection but rather

thoughtful, evidence-informed integration

emphasizing teacher agency, pedagogical

grounding, collaborative knowledge

construction, and ongoing critical evaluation.

The co-tutoring framework explored in this

study; positioning AI as collaborative partner

extending rather than replacing human

capacities, offers one promising model for

navigating this complex terrain, though

continued research across diverse contexts and

populations will be essential for developing

robust, generalizable principles.

Referencias Bibliográficas

Al Braiki, B. (2020). Artificial intelligence in

education and assessment methods. Bulletin

of Electrical Engineering and Informatics,

9(5), 1998–2007.

https://doi.org/10.11591/eei.v9i5.1984

Arini, D. (2022). Artificial intelligence (AI)-

based mobile learning in ELT for EFL

learners: The implementation and learners’

attitudes.

Canhoto, A. (2020). Artificial intelligence and

machine learning as business tools: A

framework for diagnosing value destruction

potential. Business Horizons, 63(2), 183–

193.

Celik, I. (2023). Towards Intelligent-TPACK:

An empirical study on teachers’ professional

knowledge to ethically integrate artificial

intelligence (AI)-based tools into education.

Computers in Human Behavior, 138,

107468.

https://doi.org/10.1016/j.chb.2022.107468

Chen, L. (2020). Artificial intelligence in

education: A review. IEEE Access, 8,

75264–75278.

https://doi.org/10.1109/ACCESS.2020.2988

510

Ercan, H. (2018). Pronunciation problems of

Turkish EFL learners in Northern Cyprus.

International Online Journal of Education

and Teaching, 5(4), 877–893.

Fahimirad, M. (2018). A review on application

of artificial intelligence in teaching and

learning in educational contexts.

International Journal of Learning and

Development, 8(4), 106–118.

Fernandez-Solis, J. (2013). Survey of

motivations, benefits, and implementation

challenges of Last Planner System users.

Journal of Construction Engineering and

Management, 139.

https://doi.org/10.1061/(ASCE)CO.1943-

7862.0000606

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 235

Feuerriegel, S. (2022). Bringing artificial

intelligence to business management. Nature

Machine Intelligence, 4(7), 611–613.

Gutierrez, S. (2022). Artificial intelligence in e-

learning plausible scenarios in Latin

America and new graduation competencies.

IEEE Revista Iberoamericana de

Tecnologías del Aprendizaje, 17(1), 31–40.

Haryanto, E. (2019). Students’ attitudes

towards the use of artificial intelligence SIRI

in EFL learning at one public university.

International Seminar and Annual Meeting

BKS-PTN Wilayah Barat, 1(1).

Hwang, G. (2020). Vision, challenges, roles and

research issues of artificial intelligence in

education. Computers and Education:

Artificial Intelligence, 1, 100001.

https://doi.org/10.1016/j.caeai.2020.100001

Hwang, W. (2022). The integration of multiple

recognition technologies and artificial

intelligence to facilitate EFL writing in

authentic contexts. International Conference

on Information Technology.

https://doi.org/10.1109/InCIT56086.2022.1

0067490

Jiang, R. (2022). How does artificial

intelligence empower EFL teaching and

learning nowadays? Frontiers in

Psychology, 13, 1049401.

Keezhatta, M. (2019). Understanding EFL

linguistic models through relationship

between natural language processing and

artificial intelligence applications. Arab

World English Journal, 10(4), 251–262.

Kumar, S. (2023). Artificial intelligence and

blockchain integration in business: Trends

from a bibliometric-content analysis.

Information Systems Frontiers, 25(2), 871–

896.

Lara, R. (2023). La inteligencia artificial:

análisis del presente y futuro en la educación

superior. Revista Científica Multidisciplinar

G-Nerando, 4(1).

Lu, Y. (2019). Artificial intelligence: A survey

on evolution, models, applications and future

trends. Journal of Management Analytics,

6(1), 1–29.

Nazari, N. (2021). Application of artificial

intelligence powered digital writing assistant

in higher education: Randomized controlled

trial. Heliyon, 7(5), e07014.

https://doi.org/10.1016/j.heliyon.2021.e070

Pikhart, M. (2020). Intelligent information

processing for language education: The use

of artificial intelligence in language learning

apps. Procedia Computer Science, 176,

1412–1419.

Popenici, S. (2017). Exploring the impact of

artificial intelligence on teaching and

learning in higher education. Research and

Practice in Technology Enhanced Learning,

12(1), 1–13.

Salas-Pilco, S. (2022). Artificial intelligence

applications in Latin American higher

education: A systematic review.

International Journal of Educational

Technology in Higher Education, 19(1), 1–

20.

Sumakul, D. (2022). Artificial intelligence in

EFL classrooms: Friend or foe? LEARN

Journal, 15(1), 232–256.

Sun, Z. (2021). Design of online intelligent

English teaching platform based on artificial

intelligence techniques. Computational

Intelligence, 37(3), 1166–1180.

Wang, P. (2019). On defining artificial

intelligence. Journal of Artificial General

Intelligence, 10(2), 1–37.

https://doi.org/10.2478/jagi-2019-0002

Wang, Z. (2022). Computer-assisted EFL

writing and evaluations based on artificial

intelligence. Library Hi Tech, 40(1), 80–97.

Zawacki-Richter, O. (2019). Systematic review

of research on artificial intelligence

applications in higher education – where are

Ciencia y Educación

(L-ISSN: 2790-8402 E-ISSN: 2707-3378)

Vol. 7 No. 2.2

Edición Especial II 2026

Página 236

the educators? International Journal of

Educational Technology in Higher

Education, 16(1), 39.

https://doi.org/10.1186/s41239-019-0171-0

Zhai, C. (2023). Artificial intelligence dialogue

systems for enhancing EFL students’

interactional competence. Computers and

Education: Artificial Intelligence, 4, 100134.

Zhang, J. (2018). The new developments of

constructivism theory and its reflection on

college English teaching in the era of new

media. Theory and Practice in Language

Studies, 8(6), 649.

https://doi.org/10.17507/tpls.0806.15

Esta obra está bajo una licencia de

Creative Commons Reconocimiento-No Comercial

Erika Mora Herrera, Roddy Real Roby y Gabriela

Almache Granda.