Will AI Replace Teachers? How AI Is Reshaping TEFL and ESL Jobs and Hiring

Will AI replace teachers? Not in the classroom, but it is already replacing the people who used to read your CV. Across TEFL and ESL hiring, AI now screens applications, scores recorded interviews and ranks candidates before a human looks. The teacher is still hired by a person. Getting in front of that person is what has changed.

Key takeaways

• In a June 2026 Resume Genius survey of United States hiring managers, 87 per cent said they use AI somewhere in recruitment and 58 per cent use it to screen CVs and applications. Language schools and online platforms buy the same software as everyone else.
• The UK Information Commissioner’s Office audited AI recruitment tool providers between August 2023 and May 2024 and issued 296 recommendations, including against tools that inferred a candidate’s gender or ethnicity from their name.
• Demand is shifting rather than vanishing. South Korea issued 1,286 E-2 teaching visas in the first quarter of 2026, down 47 per cent on the same period in 2020, according to Ministry of Justice figures reported by the Korea JoongAng Daily, with AI learning apps named as a cause.
• In the British Council’s study of 1,348 English teachers, 51 per cent rejected the idea that AI could teach English without a teacher by 2035, and only 20 per cent felt sufficiently trained to use it.
• Under the EU AI Act, AI used to recruit or select workers is classed as high risk. After the 2026 Digital Omnibus, those obligations apply from 2 December 2027.
• The teachers winning in this market are the ones whose applications survive the software and whose interviews show the judgement the software cannot assess.

What is actually changing in TEFL and ESL hiring?

The change is in the front half of the hiring funnel, not the back half. AI tools now source, sort, score and shortlist TEFL and ESL candidates, while the final decision still sits with a director of studies, a recruiter or a programme coordinator who is very much a human being with a coffee going cold.

That matters because for twenty years the TEFL job hunt ran on volume. You sent the same CV to forty schools, three replied, one interviewed you on Skype and you packed a suitcase. The volume game is over, because the thing reading your forty applications is now software that compares all forty of them in the same second and discards the ones that do not match the job description closely enough.

The UK government’s Responsible AI in Recruitment guidance, published by the Department for Science, Innovation and Technology in March 2024, maps AI onto four stages of hiring: sourcing, screening, interview and selection. It also records a risk that should interest anyone applying to teach English abroad. Transcription tools used on interview recordings, the guidance says, can be biased against regional and non native English speakers. The same software that filters candidates for a language school can misjudge the people most likely to apply to one.

None of this means a robot is choosing who teaches in Madrid next September. It means the robot is choosing who gets to be considered, which for the applicant is almost the same thing.

Where is AI already sitting in the hiring process?

AI sits at every stage before the human interview, and increasingly inside it. The clearest recent picture comes from a Resume Genius survey of United States hiring managers fielded in June 2026: 87 per cent use AI in recruitment, 58 per cent use it to screen CVs and applications, 44 per cent use it to match candidates to jobs on skills, 41 per cent use it to schedule interviews and 18 per cent run AI assisted video interviews. Those are general figures rather than ELT specific ones, but the online teaching platforms and international school groups that hire at scale are precisely the employers who buy this software first.

The ICO’s audit of AI recruitment tools describes what the software actually does. Screening tools score competencies from written applications and predict how likely a candidate is to succeed in selection. Selection tools score written answers to interview questions, score transcripts of video interviews, and in some cases evaluate a candidate’s language, tone and content on video to predict personality type. The ICO found tools that let recruiters filter out candidates by protected characteristics, tools that inferred gender and ethnicity from a name, and tools collecting far more personal data than they needed. It made 296 recommendations and 97 per cent were accepted.

Here is how the stages compare for a typical TEFL applicant. Figures are approximate and describe a typical range rather than any one employer.

The demo lesson row is the one to notice. Government programmes still interview people. EPIK, the English Program in Korea run under the Ministry of Education, invites successful applicants to an online interview with its own staff before any documents are requested. Private schools and platforms, which hire in far greater numbers, are the ones automating the front door.

What are employers now asking TEFL teachers for?

Employers are asking for the same core qualification as before, plus evidence that you can work with AI rather than be replaced by it. The review above is of The TEFL Institute Group’s accredited course, the group behind tefl.ai, and the phrase worth noticing is “a career in the TEFL world”. The qualification gets you through the first filter. What happens next depends on skills that did not exist on a job description five years ago.

Three things have moved up the list. The first is a verifiable certificate rather than a PDF, because document checks are now automated and an unverifiable certificate is quietly scored as no certificate. The second is a recorded presence: schools want to see you teach on camera before they fly you anywhere, and recorded demos are compared against each other. The third is AI literacy, which an increasing number of schools now treat as a professional competence in its own right.

That third point has official backing. UNESCO’s AI competency framework for teachers, published in 2024, sets out 15 competencies across five dimensions, from a human centred mindset and the ethics of AI through to AI pedagogy and AI for professional development. When UNESCO began the work, only seven countries had any AI competency framework for teachers at all. Ministries, international school groups and the larger online platforms are now writing it into recruitment criteria, which is why documented competence such as the AI-Skilled Teacher Certificate has become a sensible line on a TEFL CV rather than a novelty.

Is demand for English teachers falling because of AI?

Demand is moving, and in some markets it is falling, but the cause is a mix of AI, policy and demographics rather than AI alone. The honest answer for anyone asking whether TEFL jobs are disappearing is that certain job types are shrinking fast while others are growing, and the difference is visible in official data.

South Korea is the clearest case. The Korea JoongAng Daily reported in May 2026, using Ministry of Justice statistics, that 1,286 E-2 foreign language instructor visas were issued in the first quarter of 2026, a six year low and 47 per cent below the 2,448 issued in the same quarter of 2020. The paper linked the decline to AI learning apps that let people study without a teacher, to better instant translation changing how important English feels for a job, and to the closure of many private academies. Speak, an AI speaking practice app, was Google Play’s app of the year in Korea in 2024 and had around 316,000 active Korean users by March 2026.

China moved earlier and for different reasons. The Ministry of Education reported in December 2021 that the number of offline academic tutoring institutions had fallen by 83.8 per cent and online ones by 84.1 per cent after the Double Reduction policy. That was regulation, not AI, but it removed the online tutoring platforms that had been the largest single employer of remote ESL teachers, and AI tutors are now filling part of the gap those platforms left.

Against that, the wider market has not collapsed. The group’s own Future of AI in English Language Teaching and TEFL report estimates more than two million TEFL positions opening globally each year and some 12 million English teachers worldwide, with the fastest growth in Asia Pacific and the Middle East. The jobs are there. They are more competitive, more often filtered by software, and less often the easy conversational hagwon or cram school roles that once absorbed thousands of first year teachers.

Will AI replace teachers in the classroom?

No, and the people best placed to know are the teachers themselves. In the British Council’s study of 1,348 English language teachers across 118 countries, 51 per cent disagreed that AI could teach English without a teacher by 2035, and the most common reason given was the human element of the classroom. We looked at that dataset in detail in our piece on the future of AI in English language teaching, and the short version is that AI is excellent at the layer of teaching that happens before the lesson and poor at the layer that happens during it.

The same study carries the warning that matters for hiring. Only 20 per cent of those teachers felt sufficiently trained to use AI, while 54 per cent felt inadequately trained. Employers can see that gap too, and a growing number are hiring for it. A teacher who can show they use AI to plan, level and mark without handing over judgement is a cheaper hire to onboard than one who has never opened the tools, and a far safer one than a teacher who pastes learner data into a free chatbot.

The UNESCO framework and the ICO both point the same way: AI should complement the teacher rather than replace them, and a human must stay accountable for high stakes decisions. The EU AI Act goes further and writes it into law. AI systems used for recruitment and selection, and systems used to evaluate learning outcomes in education, are listed as high risk. Those obligations were due in August 2026 and, following the Digital Omnibus agreed in 2026, now apply from 2 December 2027. Whatever the final date, the direction is settled. Software may rank you, but a person has to be able to explain why.

How do you get past AI screening for a TEFL job?

You get past AI screening by giving the software exactly what it is scoring for and giving the human what the software cannot score. That means a CV whose wording mirrors the job advert, a certificate the checker can verify, a recorded interview that is clear and structured, and a demo lesson that shows judgement.

Start with the CV. Screening tools score against the job description, so if the advert says “young learners”, “CEFR” and “online classroom management”, those exact phrases need to appear in your application, not approximate synonyms. Our guide to writing a TEFL resume with AI walks through the drafting process, and the free CV and cover letter generator on tefl.ai produces a first version built around a specific advert in a couple of minutes. Edit it hard afterwards. In the Resume Genius survey, 82 per cent of hiring managers said they were concerned about candidates using AI in applications and 58 per cent had already received AI generated CVs or cover letters, so a generic draft is now a recognisable type of document rather than a clever shortcut.

Then the interview. Recorded one way interviews are transcribed and scored, which rewards short answers with a clear structure: the situation, what you did, what happened. Accent is not the problem, mumbling is, because transcription software handles clear speech from any variety of English far better than fast speech from any variety. The AI mock interview approach we described earlier this year is useful here precisely because it makes you practise answering a machine. Note the survey finding that 46 per cent of hiring managers have caught candidates using AI to answer interview questions live. They are looking for it, and it reads as exactly what it is.

Before any of that, check you are applying for jobs you can actually be hired for. Eligibility is still decided by passport, degree, certificate hours and background checks, as our explainer on how TEFL eligibility works sets out, and no amount of keyword tuning changes that. The free Job Readiness Checker gives you an honest read on where you stand before the software does.

What should a TEFL teacher do about this now?

Treat the hiring process as a system with two gates and prepare for both. The first gate is automated and tests fit on paper. The second is human and tests judgement in a room, real or virtual. Most unsuccessful applications in the current market fail at the first gate without the applicant ever knowing it existed.

In practical order: choose a course whose certificate can be verified online, because automated document checks are now routine, and our guide to choosing a TEFL course covers what to look for. Build a CV and a short teaching video that answer the advert directly. Learn the AI tools well enough to describe your workflow in an interview, because “I use AI to draft levelled materials and check every item before class” is now a hiring signal. Then use the Job Market Explorer to aim at markets where demand is growing rather than the ones where the visa statistics are heading the wrong way.

The teachers who will do well in the next five years are not the ones who fear AI or the ones who outsource their thinking to it. They are the ones who understand that it now sits between them and every job they apply for, and who learn to walk through it with their judgement intact.