The Regulatory and Human Challenges of EdTech: Insights from Savannah Pieters

EdTech expert Savannah Pieters analyzes AI's rapid growth, implementation barriers, and global trends shaping the future of education technology

What if the explosive growth of AI in education is being quietly tempered by a set of stubborn, human-centric challenges? While forecasts predict a hyper-acceleration of AI-EdTech, the real story lies in the gap between potential and practical implementation.

To navigate this complex landscape, we spoke with global EdTech strategy consultant Savannah Pieters, Co-Founder & Head of Marketing at Success 101, to separate the market hype from the on-the-ground realities shaping the future of learning.

Interview with Savannah Pieters

Can you quantify the current and projected growth of the global EdTech and AI-in-education markets?

The EdTech market is still scaling fast. Recent industry reports peg the global EdTech market in the low-hundreds of billions today and forecast high-teens CAGR over the next 3–5 years — with some AI-specific reports projecting even stronger growth for AI-in-education (much higher CAGR for that subsegment). In plain terms: expect the overall market to keep growing strongly and the AI-powered slice to expand much faster than the general EdTech market.

Grand View Research estimates the broader EdTech market at about $163B (2024) and projects a ~13% CAGR to 2030 (so the market roughly doubles over a multi-year horizon).

Specialist AI-in-education forecasts show much higher projected growth rates for the AI submarket (some vendors/analysts show CAGR figures in the 30–40% range across the next decade), reflecting explosive investment and adoption in adaptive learning, assessment automation, tutoring assistants and content generation. (examples: Precedence/other market briefings).

Takeaway in one line: EdTech → steady, fast growth; AI-EdTech → rapid, potentially hyper-accelerating growth that will outpace the broader market.

What are the most common barriers institutions face when trying to implement EdTech and AI solutions?

Institutions repeatedly hit the same barriers: infrastructure and digital equity, data quality & privacy, teacher capacity and buy-in, academic integrity and assessment design, and unclear governance/ethics rules. Those are the places projects stall or underdeliver.

These are some concrete challenges:

• Infrastructure & access — hardware, bandwidth, and device inequality make deployment uneven (especially outside major urban centers).

• Data privacy & governance — collecting, storing and using learner data safely (and legally) is a major obstacle; institutions worry about compliance, student safety, and consent. UNESCO and OECD guidance highlight privacy and ethics as central.

• Teacher readiness & pedagogy — teachers need training and time to redesign curricula and assessments for AI-enabled workflows; otherwise AI becomes a novelty, not a tool. Research shows many educators modify assessments because of cheating risks and to make AI useful.

• Accuracy, bias and content quality — AI outputs can be wrong, biased, or misaligned with learning goals; that forces human oversight and new QA workflows.

How does the adoption and growth of EdTech vary across major global regions?

North America currently leads in market size and depth of deployments, but Asia-Pacific is the fastest growing region (big student populations, heavy public & private investment, and hungry ed-tech markets). Europe sits between the two (strong regulation and pockets of deep adoption). LATAM is growing fast but from a smaller base.

Evidence & drivers:

• North America: largest revenue share today (advanced infrastructure, big vendors, deep venture funding).

• Asia-Pacific: repeatedly called out as the fastest-growing region for EdTech / AI-EdTech, driven by population scale, government digitization programs, and private investment. Reports show APAC posting the highest CAGRs.

• Europe: adoption is steady but shaped heavily by data protection rules and stronger regulatory scrutiny (which can slow rollouts but raise quality).

• LATAM: strong interest and growing pilots; growth is accelerating with major regional investments but still smaller absolute market vs NA / APAC. Recent MNE investments and partnerships (cloud, AI) are ramping up capacity in LATAM.

APAC—fastest growth; North America—largest current market; Europe—regulated but strong; LATAM—emerging growth.

How are governments and regulatory bodies beginning to shape the deployment of AI in education?

Governments and supranational bodies are waking up fast: they’re releasing guidance, starting consultations, and pushing frameworks that force institutions to plan for privacy, fairness and safety before they deploy AI at scale. That means regulators are moving from “wait and see” to “guardrails first.”

What they’re doing:

• UNESCO and other agencies have issued guidance aimed at policymakers on how to govern AI in education — focusing on transparency, human oversight, data protection, equity, and teacher support.

• OECD / national regulators: providing frameworks and studies to help ministries assess tradeoffs; many countries are drafting educational-AI policies addressing data protection and ethical use.

Practical result: Institutions are being asked (and in many cases required) to document data flows, obtain informed consent, apply de-identification, and set teacher/human oversight protocols — all before wider rollouts.

Implication: regulatory pressure raises the bar (good), but slows some deployments — so compliance is now a strategic implementation task, not an afterthought.

What is the realistic impact and measurable return on investment for generative AI tools in educational settings today?

Generative models are already useful — for personalised explanations, content generation, feedback, and tutoring scenarios. They can increase engagement and speed up content creation. But they’re not a magic bullet: accuracy problems, hallucinations, fairness/bias risks, and weak long-term learning evidence in some contexts mean humans must stay in the loop.

Evidence & nuance:

• Benefits observed: personalized explanations, rapid content/material generation, formative feedback and scalable tutoring have improved engagement and sometimes short-term scores in pilots and studies. Several SLRs and reports show promising signs in higher education and corporate L&D.

• Limitations: Innacuracy (incorrect answers), inconsistent depth of reasoning, potential bias, and content quality variance — plus concerns about students over-relying on AI and missing skill development. Research highlights both gains and real concerns about integrity and accuracy.

ROI is usually measured in a mix of learning outcome metrics (completion, test scores), operational efficiency (teacher time saved, grading automation), engagement (time on task, retention) and business metrics (time-to-competency, revenue uplift for corporate training). The most impactful, measurable use cases are onboarding/upskilling in corporate L&D, automated assessment/feedback in scale learning, and adaptive practice that increases mastery rates.