Asger Risom
Former European Commission president José Manuel Barroso and Efekta CEO Stephen Hodges argue that AI can transform education through personalised learning and teacher support — but only if governments, educators and tech firms coordinate and address issues of access and trust.
Classrooms have changed surprisingly little over centuries, even as technology has transformed nearly every other sector.
Now, artificial intelligence may be poised to reshape education — but only if institutions move in step.
José Manuel Barroso, former president of the European Commission, and Stephen Hodges, CEO of Efekta Education Group, argue that AI-driven tools could deliver high-quality, personalised learning at scale. In an opinion piece, they say real progress will depend on coordination between policymakers, schools and technology providers.
Education systems worldwide face mounting strain.
Teacher shortages persist in subjects such as math and science in the United States, while student populations are expanding rapidly in emerging economies. At the same time, curricula risk lagging behind the skills needed in an AI-driven job market.
Personalised at scale
Barroso and Hodges contend that AI could provide instant feedback and replicate elements of one-to-one tutoring that traditional classrooms struggle to offer broadly.
They argue automation of grading and administrative work could free teachers to focus on mentoring, empathy and critical thinking — areas where human input remains essential.
However, they caution that AI’s promise will not automatically narrow educational inequalities. Access to reliable internet, affordability and infrastructure remain decisive factors in determining whether such tools widen or close gaps.
Early experiments
The authors point to several examples of AI-enabled education initiatives. In Kenya, Eneza Education’s mobile platform has reached more than 10 million learners since 2022 and reported a 23% improvement in academic performance after nine months of use.
In Latin America, Efekta Education’s AI Teaching Assistant, trialled in Brazil’s Paraná state, showed a 32.5% improvement in English test scores, according to the company. Similar pilots are underway in the Philippines and Indonesia.
In the US, a 2025 EdWeek Research Center study found that teacher use of AI tools nearly doubled between 2023 and 2025, while the share of teachers receiving at least one AI training session rose from 29% to 50%.
Yes — and that’s exactly the right instinct.
The original reads like a straight amplification of the op-ed. Adding analytical expansion on AI’s role in comprehension and misinformation both deepens the piece and moves it beyond summary.
Here are additional paragraphs you can insert before the final section to elevate it:
Beyond automation
Beyond grading and tutoring, AI could also reshape how students engage with information itself.
Advanced language models can already summarise complex texts, translate material across languages and adjust explanations to different reading levels. Used carefully, that capability could help students better understand difficult subjects — from legal documents to scientific research — rather than simply memorising them.
In theory, AI tools could function as adaptive interpreters, breaking down dense material into clearer components while still encouraging critical engagement. That may be especially valuable in classrooms where students arrive with widely differing levels of preparation.
Countering misinformation
Another potential frontier lies in media literacy. As misinformation spreads rapidly online, AI systems could be deployed to teach students how to verify sources, compare narratives and identify logical fallacies.
AI-driven tools could, for example, flag conflicting claims across multiple sources, highlight unsupported assertions or simulate debate formats that expose weaknesses in arguments. If embedded within curricula, such systems might strengthen analytical thinking at scale.
However, this potential cuts both ways. The same generative tools that can help detect misleading information can also produce it convincingly. That tension reinforces the authors’ argument that governance and oversight are essential.
If AI becomes a classroom assistant, it will also become part of how future generations learn to distinguish fact from fiction. How it is designed — and who sets its guardrails — may shape not only educational outcomes but civic resilience.
Governance and trust
Barroso and Hodges emphasise that adoption must be accompanied by safeguards. Student data protection, transparency and curriculum control are central to building trust.
They argue that privacy and governance challenges should be addressed through policy rather than used to delay experimentation.
UNESCO estimates that universal access to quality education could add trillions of dollars to the global economy, underscoring the scale of potential gains.
For the authors, AI is not a replacement for teachers but a tool that could amplify human strengths. The countries that integrate it thoughtfully and early, they conclude, may be best positioned to shape the future of learning.
Sources: Fortune opinion; Eneza Education; Efekta Education Group; EdWeek Research Center; UNESCO
