In December 2024, a hacker breached the PowerSchool software platform. This platform stores student information for schools across the US. The hacker didn’t need any special skills. They just used one stolen password and got straight in.
By January 2025, the hacker had pulled 62 million student records from that system: names, Social Security numbers, medical histories, and accommodation notes – children’s data. The portal didn’t even require multi-factor authentication.
That happened six months ago. It’s still the clearest illustration of what’s at stake when EdTech organizations treat compliance as the finish line rather than the starting point.
Researchers project the AI in the education market to grow from $6.9 billion today to $41 billion by 2030. Tutoring businesses are moving fast on AI: scheduling, student progress tracking, and learning recommendations. Most tutoring businesses jump into using AI tools first and ask questions later. That gap is where problems start.
The 4 Ethical Concerns of AI in Education Every Tutoring Business Needs to Own
1. Data Privacy: Compliance Isn’t the Same as Ethics
Laws like FERPA, GDPR, and the UK’s Data Protection Act? They set a floor. Not a ceiling. Staying compliant means you haven’t broken the law. It doesn’t mean you’ve stopped to ask what data you actually need from minors, how long you’re keeping it, and what happens when something goes sideways.
PowerSchool skipped multi-factor authentication. That’s it. One missing step that the rest of the industry figured out a long time ago. Nobody had enforced a data retention policy that matched the actual risk, so records dating back to 1985 sat exposed. Compliant on paper. Careless in practice.
Before signing with any AI platform, get these answers in writing:
- Where does the vendor store student data?
- Does the vendor use student data to train their models?
- Who at that company can access individual student records?
Vague answers are themselves an answer. Wise.Live builds its platform around a single idea. Efficiency and data responsibility can coexist. The scheduling, invoicing, and progress reporting tools cut admin load for tutors without treating student data as a secondary revenue stream.
2. Algorithmic Bias in Education: The Ofqual Lesson
In August 2020, the UK government canceled A-level exams and tasked Ofqual with deploying an algorithm to predict grades. The logic: anchor each prediction to the school’s historical performance. What followed became one of the most visible examples of algorithmic bias in education in recent years:
- Ofqual downgraded 39% of predicted grades, hitting state school students hardest
- Private schools recorded a 4.7 percentage-point rise in top grades, more than double what comprehensives saw
- Hundreds took to the streets outside the Department for Education
- Ofqual’s chief regulator quit within days. Universities pulled offers
The algorithm did exactly what developers built it to do. That was the problem. They trained it on decades of unequal data and sent it live without a single bias check.
Every AI tool learns from its training data. Bias goes in, bias comes out. Before signing with any vendor, ask two things: what did you train this on, and who checked it for bias?
3. Generative AI Is Breaking Academic Integrity. Detection Isn’t the Fix.
Universities formally caught nearly 7,000 UK students using AI to cheat in 2023-24, three times the figure from the year before. By 2025, 88% of UK students had used generative AI in assessments, up from 53% the previous year.
Detection tools haven’t fixed it. Australia’s regulator TEQSA warned that AI cheating is “all but impossible” to detect consistently. Detection software also flags non-native English speakers and non-standard writing styles, adding an equity problem to an integrity one.
The real question isn’t “did this student use AI?” It’s “did they actually learn anything?” Two students, two very different outcomes:
- One uses AI to get past a writing block, then critically reworks the output. That student builds judgment.
- One pastes a prompt and submits it unchanged. That student bypasses the cognitive work entirely.
The only person who can tell the difference is a tutor who genuinely knows how that student thinks. No detection tool gets close to that. That’s why a strong tutoring session structure outlasts any detection algorithm when it comes to protecting academic integrity.
4. Over-Reliance and the Slow Erosion of Critical Thinking
Nobody talks about how AI affects critical thinking. That’s a mistake.
A 2025 study across 666 participants found that students who used AI most frequently scored lowest on critical thinking assessments. Students pass the thinking to the tool.
Over time, they stop generating their own reasoning altogether. A Centre for Democracy and Technology report found 70% of teachers worry AI weakens students’ critical thinking, even in a year when 85% of teachers and 86% of students used AI regularly.
AI optimizes for efficiency. Education isn’t an efficiency problem. The productive struggle, the retrieval practice, the process of being wrong and working out why, that’s where learning actually happens. AI short-circuits all of that when students use it as a substitute rather than a scaffold.
Tools that cut admin load free tutors to do what AI genuinely can’t: build the relationship that underpins real learning. That’s the clearest expression of the ethical use of AI in education in day-to-day tutoring. Wise.Live’s student progress tracking tools surface data for tutors to interpret. Tutors decide. The cognitive work stays human.
Accountability: Who Answers When AI Gets It Wrong?
A 2025 systematic review in Nature identified “absence of accountability” as one of the three primary societal risks of AI in education. Most educational AI gives you outputs with no explanation. A student gets a bad recommendation, and no one can trace it back to its source.
- Your parents don’t know how AI fits into their child’s sessions. Change that.
- Audit tools for bias and accuracy, not just how well they perform.
- When a decision matters, a person makes it. Not a dashboard.
- Tutors need to know where AI fails, not just where it works.
AI ethics in education isn’t something you write down once. Your team either practices it every week or they don’t.
Why the Most Responsible Tutoring Businesses Will Win
Speed won’t separate the leaders from the rest. Judgment will.
The tutoring businesses that earn lasting trust over the next five years won’t be the ones that adopted every AI tool first. They’ll be the ones who clearly explain what they use, why they use it, and who answers when something goes wrong. Parent trust is an undervalued differentiator in this market, and right now, most of the competition isn’t even thinking about it.
The ethics of AI in education are what parents are beginning to ask about. The businesses that already have answers will be the easiest to choose.
Frequently Asked Questions
What is the ethical use of AI in education?
At its core, ethical AI in education means humans stay in charge. Data gets protected, bias gets caught, parents stay informed, and every decision that matters has a person behind it.
What do tutoring businesses actually need to worry about when it comes to AI?
Four things. Student data falling into the wrong hands. AI tools quietly reproduce bias at scale. Students are submitting AI output without actually learning anything. Critical thinking is slowly eroding as more students lean on AI to do their thinking for them.
What is algorithmic bias in education?
Ofqual’s 2020 grading scandal said it all. One algorithm, trained on school history, handed private school students better grades while state school students lost university places. That’s algorithmic bias in education: AI inheriting the inequalities already sitting in the data.
Who is responsible when AI gets an educational recommendation wrong?
Everyone points at someone else. Developers blame the data. Businesses blame the vendor. But regulation is shifting fast, where the organization that deploys the AI now carries the liability, not just the one that built it. Vendor vetting and human oversight aren’t optional extras anymore.
How can tutoring businesses use AI ethically?
Vet vendors on data governance before signing anything. Parents deserve to know how AI touches their child’s learning. Audit your tools for bias, not just results. And when a decision matters, a human makes it. AI feeds the thinking. People own it.
