Why Edtech is core infrastructure for the AI economy

AI is eating the world - but without widespread education, the gains will be uneven, fragile, and ultimately constrained. 

At Brighteye, we’re betting on the diffusion curve, not just the frontier models. That’s why, in our opinion, Edtech is the most overlooked lever in unlocking AI’s true societal and economic value.

AI literacy is the new digital literacy and a pre-cursor for (responsible) mass use. We’ve reached a point where every knowledge worker, frontline employee, and small business operator needs to understand AI - not just how it works, but how to use it responsibly and effectively. 

This can’t be solved with a few YouTube tutorials or corporate workshops. It requires infrastructure. It requires scale. It requires smart, adaptive education platforms that are focused on specific use cases that help us collectively maximise its potential. 

Sweden, as of last week, is taking a bet to get ahead of the curve. Its AI reform, led by Sana in partnership with the government, is a first-of-its-kind national strategy to upskill the government workforce in AI. It’s happening now, using AI agents to deliver personalised, high-quality training, administrative efficiency and improved decision-making for public sector workers from nurses to teachers. This is what systemic adoption looks like.

Image lifted from Sana's promotional video

We spoke with Jon Lexa, Sana's President - he emphasised the importance of public-private partnerships:

"We believe that public and private sectors will have to work together to drive AI literacy at scale and we collectively have to act today, not tomorrow - making bold bets now will enable competitiveness on an international scale."

As Sana itself says, “AI is only as good as the context it’s given. With Swedish AI Reform, Sana empowers civic organisations to build AI agents that understand and act on their institution’s data–helping them find information, automate tasks, and solve complex problems.”

The implications are obvious: AI will touch every sector, but without AI literacy, those gains stall. We won’t get the productivity boost, the improved services, or the responsible oversight unless people actually understand the technology. That’s both the bottleneck and the opportunity.

Y Combinator’s latest Request for Startups captures the moment perfectly: Edtech tools that help people use AI to become more productive are squarely in focus. This isn’t traditional “Edtech” in the K12 or MOOCs sense. This is about arming the workforce with capabilities and doing so as quickly as possible. It’s about tooling up professionals to compete, adapt, and innovate in a world where AI is becoming foundational.

Modern Edtech platforms aren’t just delivery vehicles, they are applications of AI. The best ones offer real-time adaptation, contextual learning, and frictionless onboarding. The unit economics are strong and the TAM is clearly massive. And the demand, both at the individual and enterprise level, is only growing.

The societal upside is equally compelling. When people understand AI, they use it better. They ask smarter questions, build more ethical systems, and push for transparency. It’s not just about technical training; it’s about cultivating a generation that can think critically about AI’s role in work, policy, and daily life.

Investing in AI without investing in AI literacy is short-sighted. It’s like building a new internet without teaching anyone to use a browser. If we want real impact, we need to fund the tools that make AI usable and useful for everyone.

Edtech is no longer an adjacent category. It’s core infrastructure for the AI economy.

—— 

To make this point, we wrap with the sub-categories called out in Sana’s announcement:

1. Transforming education

   
   

Students need to learn how to use agentic AI effectively and responsibly. We want to make Sana Agents available to students aged 13 and up while ensuring it is conducive to learning, and equip educators with the tools they need to reimagine education for an AI-first world.

• Student AI literacy: School and university students can develop the critical skills needed to leverage AI in their future careers.

• Curricula innovation: Teachers and program directors will be able to re-design programs and experiment with new teaching methods.

• Administrative operations: Educators can spend more time on teaching and mentorship with AI supporting in grading, reporting, and student onboarding.

2. Powering public good

Government institutions that embrace agentic AI won’t just improve efficiency—they’ll unlock new ways of serving their citizens. By equipping civil servants with the latest AI tools, we can build more responsive, data-driven institutions that solve problems before they arise.

• Urban infrastructure: Agencies can develop more responsive cities by predicting maintenance needs and optimising resource distribution.

• Administrative operations: Civil servants will be able to shift their focus from managing bureaucracy to driving meaningful change with AI automating repetitive tasks like citizen enquiries and permit processing.

3. Unlocking social impact

Non-profits do crucial work with limited resources. AI agents grounded in organisational data can handle operations behind the scenes so staff can spend more time on advancing their missions.

• Donor engagement: Fundraisers can strengthen donor relationships with AI analysing giving patterns, automating outreach strategies, and personalising campaigns based on engagement history.

• Task automation: Staff will reclaim valuable time for mission critical work by handing over repetitive, administrative tasks like gift processing and grant management.

• Program development: Non-profit leaders can amplify their organisation’s impact by leveraging AI to analyse program effectiveness and identify new funding opportunities.

4. Advancing research breakthroughs

Despite global research effort rising substantially, research productivity is in sharp decline. We want Sweden’s scientists and research labs to increase their chances of finding the next Nobel prize-winning invention or cure.

• Combinatorial discoveries: Scientists can use agentic AI to analogise across domains and break the siloed nature of scientific research.

• Data analysis: Researchers will process vast datasets in seconds to uncover patterns and insights that might otherwise be missed.

• Research operations: Institutions can increase focus on innovation by offloading administrative tasks like drafting grant proposals and managing literature reviews.