Knowledge Hub Β· Article 177 Β· Parents

The future of online tutoring: what changes by 2030 (AI as the practice partner, the shared whiteboard as standard, recording by default, prices converging across borders, verified tutors), what does not change (one tutor, one child, a trained teacher, the school's curriculum, months not weeks), what it means for a parent choosing a tutor now, and what to be sceptical of

0min read0students taught0countries
Article 177 Β· Parents

Every tutoring company has a page about the future of online tutoring, and all of them predict a future in which that company's product is the future. This page tries to do something more useful: say what is actually changing, based on what has changed in the five years since tutoring moved online, say what the research on tutoring tells us will not change, and say what a parent should do about it now.

Five things that are changing

1. AI is becoming the practice partner between classes. The tools now generate practice questions on a topic, mark them, show a worked example, and answer a factual question at eleven at night. Used well, they do what a workbook used to do, faster and with feedback. By 2030 the normal pattern will be a tutor who teaches and diagnoses in the class and sets practice that an AI tool runs and reports on between classes. What AI is not doing, and the studies so far do not show it doing, is replacing the thing a tutor is for, which is below.

2. The shared whiteboard is replacing the plain video call. In 2020 most online tutoring was a video call with the child holding paper up to the camera. The services that produce results have moved to a shared whiteboard on which the child writes and the tutor sees every line as it appears. This is the single largest quality difference between online services now, and by 2030 a service without it will look like one without video.

3. Recording is becoming the default. Every class recorded, the recording sent to the parent and kept by the child for revision. It is a safety feature, a learning feature and an accountability feature at once, and parents are starting to ask for it as a matter of course. The rules a good service follows are in is online tutoring safe.

4. Prices are converging across borders. A qualified teacher with a master's degree in India can teach a child in Houston, Manchester or Sydney the child's own curriculum, with the same whiteboard and recording, for about a quarter of what a local teacher charges. As parents learn to check that the tutor teaches their curriculum, this is becoming the ordinary way for a family to afford weekly tutoring over months, which is the pattern the research supports. What it costs country by country, and what each price buys, is in is online tutoring worth it.

5. Verification is replacing claims. Marketplaces check identity, not teaching, and the complaints on review sites are mostly about tutors who were not what their profile said. The direction is toward services that show the tutor's degree, name the tutor, keep the same tutor, and publish prices. Parents are asking the twelve questions in how to choose an online tutor, and the services that cannot answer them are being found out.

Four things that do not change

The research on tutoring is unusually consistent about what produces the large effects, and none of it is technology.

  1. One tutor with one child, or at most two or three. The effect shrinks as the group grows.
  2. A trained teacher, not a helper, a student or an app.
  3. The child's own curriculum, in the order the school teaches it.
  4. Regular sessions over months, not a few weeks before a test.

A service that has these in 2030 will work; one that has every new feature and not these will not. The reasons, with the evidence, are in does online tutoring work.

Where AI genuinely helps, and where it cannot replace a teacher

AI does this well now A teacher is still needed for this
Generating practice questions on a named topic at a named level Deciding which topic, at which level, this particular child needs today
Marking closed-answer questions instantly Reading a child's working and finding the step that goes wrong
Explaining a worked example a second and third time, patiently Noticing that the child is nodding without understanding
Answering a factual question at any hour Judging whether the child is tired, frightened, guessing or bored, and changing the class
Keeping a log of what was practised Building the relationship that makes a child willing to be wrong, which is where learning happens
Drilling vocabulary, facts, tables Teaching a method so that it transfers to a problem the child has not seen

The honest summary: AI is a very good workbook with a patient voice. The thing that moves a struggling child is a person who sees them, and the research on AI tutoring so far shows gains in practice volume and engagement, not a replacement for that person.

What it means for a parent choosing a tutor in 2026

  • Choose on the four things that do not change, and on the whiteboard and the recording.
  • Treat AI practice tools as a bonus between classes, not as a reason to choose a service, and certainly not as a substitute for the class.
  • Expect to see the tutor's name and degree, and to keep the same tutor. If a service cannot promise that, it is a marketplace, whatever it calls itself.
  • Expect prices to reflect the teacher's qualification and country, not your postcode, and ask what the price includes (recording, notes, marked homework).
  • Pay monthly. The yearly package is the business model of the services that are losing ground, and it is where the refund complaints come from.

What to be sceptical of

  • A promised score or grade. Nobody honest promises one; what a good service promises is a diagnosis, a plan and visible change within a month.
  • "AI-powered personalised learning" as the headline. Ask who the tutor is.
  • "Personalised" group classes. A group of eight is a small classroom.
  • A tutor pool. A child cannot build a working relationship with a pool.
  • A video call without a whiteboard. The tutor cannot see the working, which is most of the point.

What we think the hour after school looks like in 2030

A child at a laptop, a tutor with a master's degree on the screen, a shared whiteboard between them with the child's working on it, the recording going to the parent afterwards, and a practice set the AI runs on Wednesday and reports on before Thursday's class. The same tutor every week from September to June. The classroom will not have changed much by then; this hour already has, and it is available now to any family that asks the right questions. Which child, subject and age suit online, and which still suit in person, is in online tutoring vs in-person.

Where to get help

A free 30-minute 1-on-1 class with a tutor who holds a master's in the subject, on a shared whiteboard, recorded, with you in the room: the form of tutoring the research supports, as it is now rather than as it is promised. The same tutor every week, monthly billing, the child's own curriculum. How a class is built is on the 1 on 1 online tutoring page.

Questions parents ask

1What is the future of online tutoring?
More of the form that works and less of the form that does not. One tutor with one child on a shared whiteboard, recorded, with a qualified teacher who teaches the child's own curriculum, at a price that qualified teachers in other countries can offer: that is growing. Generic group classes sold as tutoring, and video calls where the tutor cannot see the child's working, are losing ground. AI is becoming the practice partner between classes, not the teacher in them.
2Will AI replace online tutors?
Not for the thing tutoring is for. AI tools already generate practice questions, mark them, explain a worked example and answer a factual question at midnight, and those are useful between classes. What they cannot do is watch a particular child's working, notice the one step that goes wrong, judge whether the child is tired or frightened or guessing, and build the relationship that makes a child willing to be wrong in front of someone. Tutoring's effect in the research comes from that, and the research on AI tutoring so far shows gains in practice, not in replacing a teacher.
3Is online tutoring growing?
Yes, in every country we teach in. The move online in 2020 made it normal; what has kept it growing since is price and access: a family in a small town or abroad can have a teacher who knows their exact exam board, and qualified teachers in India can teach a child in Texas or Melbourne at a fraction of local rates, with the same recording and the same whiteboard. The market is also consolidating around the form the research supports, which is a good thing for parents.
4What will online tutoring look like in 2030?
Much like the best of it looks now, more widely available: a shared whiteboard with the tutor seeing every line, every class recorded for the parent and the child, an AI practice partner that drills what the tutor set and reports back, tutors whose degrees and teaching are verified rather than claimed, and prices that reflect the teacher's qualification rather than the parent's postcode. The classroom will not have changed much; the hour after school will have.
5How should parents choose a tutor given all this?
By the four things that do not change: one tutor with one child, a trained teacher of the child's own curriculum, regular sessions over months, and for online, a shared whiteboard and a recording. Treat AI features as a bonus for practice, not as a reason to choose a service. Be sceptical of anything that promises a score, anything sold as a yearly package, and anything that calls a group of eight "personalised".

See how we teach this, 1-on-1 online β†’

written by

Shobha

Founder and lead tutor, Science with Shobha

Teaching maths, science and English to children online since 2018. 500+ students, 20+ tutors, families in nine countries.

About the team β†’