A few thoughts on Delmar Larsen's "National Priority for Open Education"

Open education licenses are global, so the infrastructure that open resources run on needs to be global too. Infrastructure that only works where internet and power are reliable, or that's designed for one particular country, contradicts open licenses.

Delmar Larsen's call for a National Open Courseware Ecosystem gets one thing right: open education needs to become an OER ecosystem, not another collection of repositories. He correctly points out that the systems that make open content usable at scale, including homework and assessment, gradebook integration, authoring, analytics, interoperability, accessibility, AI, and sustained technical and teacher support, also need to be open.

But his national, higher-ed frame misses an opportunity. K-12 is the real test, and it's global. A system centered on North American higher education starts with a population that's already relatively well connected and surrounded by digital infrastructure. That's not where the hardest design problems are, and it's not where we should start building a new OER ecosystem. K-12 is where foundational digital skills develop, where teacher capacity matters most, and where access decisions shape a learner's entire trajectory. It's also where the world's largest population of learners is found, from underfunded American districts to classrooms in Tamale or Manila. Solving the hard problems is what generates funding.

The design test

The design test can't be: will this work in a well-connected university? It has to be: can this work in an elementary classroom with no internet? If it works without internet, the system scales upward into better-resourced environments. If it doesn't, it isn't really an open system, and it isn't very usable for the largest group of learners.

There's a practical reason to frame the project nationally: national priorities attract national money. The NSF, the Department of Education, state systems and their legislators come with identifiable institutions, constituencies, and budgets. "Global open courseware infrastructure" doesn't come with an obvious state or federal line item.

That makes a national strategy understandable; it doesn't make it the right place to start. If the funding source quietly determines the user population, the ecosystem gets optimized for the institutions that finance and govern it, while everyone else remains an afterthought. Open education can't confuse the geography of its funding with the geography of its users.

Delmar's proposed ecosystem assumes cloud-hosted platforms, real-time analytics, streaming content, and continuous integration among services. But much of the world, and increasingly much of U.S. K-12, doesn't operate under those conditions. Electricity may be intermittent. Broadband may be expensive, unreliable, or absent. Devices may be shared among students or across a whole classroom. Governors are talking about getting rid of anything that uses a screen, except when it comes to testing.

Offline-first can't be a feature bolted on afterward, because it changes the system's fundamental architecture. Connected-first systems prioritize cloud services and always-on integrations; offline-first systems require local storage, asynchronous workflows, low-bandwidth content, and the ability to operate independently of the cloud. The two sets of priorities aren't compatible, though offline-first doesn't rule out connectivity where it exists.

The historical pattern in ed-tech is to build for the well-resourced environment, then treat everyone else as a special case. A better approach designs for the harder problem first. An offline-first system still works beautifully when broadband and electricity are abundant; a cloud-dependent system can't reliably move the other direction. Designing for scarcity isn't designing for less: it's designing for resilience.

Capacity-building is infrastructure

The same principle applies to people. Teacher training, digital literacy, implementation support, and ongoing capacity development aren't things that follow the technology: they're part of the infrastructure itself. A sophisticated platform only people with existing technical skills can use doesn't democratize access; it just compounds the hardware barrier with an expertise barrier.

Capacity-building has to be designed, funded, and staffed alongside the technical architecture, and it has to be global from the start. A U.S.-centric training pipeline doesn't represent the teachers who'll use the system elsewhere. The people building and supporting it need to learn from classrooms facing the greatest constraints, not from the institutions with the largest budgets. The neighborhoods around U.S. universities are not good testing environments for ed-tech.

ADAPT shows why this matters

LibreTexts' ADAPT integrates WebWork, IMathAS, H5P, and native QTI into a single homework platform, but it doesn't work in a classroom with no internet. QTI and H5P can port into Moodle relatively cleanly, but WebWork and IMathAS depend on server-side infrastructure for randomization and grading, and features like Learning Trees, analytics, and clicker integration have no straightforward offline equivalent. This isn't a criticism of ADAPT's engineering; it illustrates the consequences of its design assumptions. Moving a platform built for connected higher ed into an offline K-12 classroom isn't a matter of changing settings, it requires architectural change. Without a radically different approach, a new ecosystem is very likely to stay stuck in the current model.

Open content hasn't displaced the infrastructure underneath it

In the current model, an open textbook can replace a commercial one, but the learning still depends on Pearson's, Cengage's, or McGraw Hill's proprietary assessment and grading systems. The textbook is free; the infrastructure that enables the learning isn't. Commercial platforms still largely control assessment, grading, analytics, and increasingly the data students generate. In K-12, Google controls most things; Google is "kinda free," but definitely not open.

The goal isn't more open content; it's open learning infrastructure, where content, assessment, data, and delivery can operate without a proprietary layer underneath, whether the classroom has gigabit broadband or none. Otherwise open education just reproduces the dependency it set out to remove: the vendors change, but the dependency remains.

This is where licensing choice stops being a side issue. A CC BY license permits commercial reuse, which sounds generous but leaves the door open for a for-profit company to repackage openly licensed content behind a proprietary platform, so the ecosystem's usability once again depends on a vendor. CC BY NC closes that door. It keeps the content free to adapt, remix, and redistribute, while blocking a commercial actor from enclosing it inside a paywalled homework system or subscription analytics dashboard.

A non-commercial license means the infrastructure built around the content (authoring tools, assessment engines, offline sync layers) stays available to the public institutions, NGOs, and ministries of education that actually need to deploy it in low-resource classrooms, rather than becoming a licensing fee for a startup that spots an underserved market. It also protects the incentive to build shared, global infrastructure at all: if any contributor's improvements can be commercially captured and resold, the case for pooling effort into one open system weakens. CC BY-NC keeps the value inside the commons.

I'm not arguing against sustainability funding, paid support contracts, or vendors building compatible tools on top of an open core for those who can afford it. I'm arguing that the core itself, the content and the infrastructure layered directly on it, has to stay open, free, and safe from being quietly re-enclosed the moment it becomes valuable enough to enclose.

The opportunity is bigger than open courseware

The opportunity isn't a better repository of open courses for U.S. higher ed. It's a genuinely open learning infrastructure that can serve a classroom anywhere: global by design; offline-first, not offline eventually; K-20-centered, not higher-ed-centered; open beyond content, including assessment and analytics; capacity-building treated as infrastructure, not an afterthought; and locally adaptable in language and curriculum.

National funding, state systems, universities, and foundations can all contribute as they're able and willing, but none of those funding sources should define who the system is for. The assumptions built into the architecture on day one determine who can easily use it. A national system that might eventually go global will carry assumptions that require expensive retrofitting later; a global, offline-first, K-20-centered system can be adopted nationally just as easily, and has far larger reach, and is more likely to attract the necessary funding.

OER already solved the problem of crossing borders; its licenses do that automatically. Now, the infrastructure needs to catch up.

Should Schools Run Their Own AI?

For most districts, using a commercial AI tool is the easy answer. But, what if we ran the AI ourselves? Hospitals, banks, law firms, and government agencies have already gone down this road, not because they love managing servers, but because they handle sensitive information and want control over where it goes and who touches it.

Every time a school district sends  student or staff  generated data to an outside AI vendor it is trusting the vendor’s contracts, security, and data practices.  Self-hosting offers another path: the school, coop or state keeps control of the data, the model, and the infrastructure itself.

This is easier to do today than it would have been even two years ago. Tools like Ollama and vLLM let organizations run capable open-weight and even open source AI models on their own hardware. The auditing and access controls that a school system would need are also available. None of this requires building AI from scratch anymore.

No school in Minnesota has tried this, yet, as far as I know, but Merced County's education office in California has already piloted self-hosted AI tools across departments serving 20 school districts, using Ollama and Open WebUI. It's a small-scale proof of concept, not proof that this works at classroom scale or saves money long-term. But it shows an education agency can actually do this. Minnesota might just be good enough and smart enough to do this, and doggone it, people like us.

Owning the infrastructure also changes how teachers and students feel about the tool. Much of today's push to be "more human" in the classroom is really a response to commercial AI and pretty much all of the other apps used in schools being controlled by a distant company, not anyone in the building. That distance breeds distrust. When a school controls its own AI server, the relationship shifts: teachers can see what the system is doing, students can ask how it actually works, and the tool feels like something the school uses on its own terms. Students and teachers might even try to do some retraining of the model. The unease about AI in schools may have less to do with AI itself than with who holds the keys.

Self-hosting doesn't have to mean every district buys its own servers. Germany offers a different model: a shared, government-coordinated AI system called BildungsLLM, built on Llama 3.3 and run through a national cloud provider, all within German data centers under GDPR rules. Individual schools don't run anything themselves; they just use the shared service. Another plus is that BildungsLLM (and its portal frontend AIS.chat) is integrated into a school's Moodle system, so the data flow remains strictly contained within a closed, secure network sovereign ecosystem

That's an option worth considering. I don’t expect many Minnesota districts will actually use Moodle again, but a large district might run its own AI system. A cluster of small districts might rely on a state or regional consortium instead. And there's a third option that is also possible: a small, air-gapped AI server, disconnected from the internet, sitting in a single school or classroom. It wouldn't replace a district's main system, but it could be a low-stakes way to teach how AI actually works, what a model is, how it responds to prompts, why it sometimes gets things wrong, all without any data ever leaving the building. For schools wary of handing data to any outside company, an air-gapped box is a way to start teaching AI literacy with almost no risk. It’s a great tool for experimenting and teaching students about AI.

Self-hosting requires hardware, electricity, security, and staff time. A small district with light AI use may still find a commercial subscription cheaper. But for a large district or service cooperative with thousands of users, the calculation is different. The money goes toward infrastructure the school system actually owns, instead of writing increasingly larger checks to a vendor indefinitely.

K-12 also gets to go second. Healthcare, finance, and government already worked out the hard parts of running AI responsibly, including plenty of costly mistakes. Schools can borrow what worked without repeating what didn't. This doesn’t mean self-hosting is right for everyone. Small and rural districts often lack the IT staff to run a self-hosted AI safely, and commercial vendors offer support contracts, uptime guarantees, and someone else absorbing the operational risk. There are trade-offs.

The real question isn't whether every school should run its own AI servers. It's whether co-ops, and large districts will build enough of their own AI infrastructure and know-how to keep control over the increasing technology, data, and costs their schools will depend on more each year.
  

Why an Offline-First Learning System Can't Be Agnostic About Content

 Imagine a hospital built for a disaster zone. Its engineers have solved extraordinarily hard problems: a power grid that holds when the regional grid fails, a water system that stays clean when supply lines are cut, redundant generators, backup batteries, and careful load management so the whole facility does not collapse the first time demand spikes. It is, by every infrastructural measure, an excellent hospital. Now imagine that same hospital's leadership says, when asked what medicine is stocked on its shelves, "That's not really our department. We built the systems that keep medicine cold and available. What medicine ends up in the building is a pharmaceutical question, not a hospital infrastructure question."

This is an absurd position for a hospital to take (although, maybe not from a U.S. healthcare system perspective), and it is absurd for a specific reason: the power grid and the water system were never the point. They were built in service of the medicine. A hospital with flawless infrastructure and an empty or arbitrary pharmacy has not achieved neutrality. It has failed at the very thing the infrastructure was meant to support.

This is the position Offline-First Design for Learning (OFDL) occupies with respect to educational content.

OFDL, developed by Marc Alan Sperber through work in refugee camps and other low-resource, crisis-affected contexts, solves real problems with real rigor. It calls for servers that run on under twenty watts, sustained by solar panels and battery packs, so that learning continues through power outages. It insists that all course materials, including videos, handouts, and articles, be integrated directly into the learning management system rather than dependent on external links, because a link to the outside world is a link to a connection that may not exist. It specifies open standards so that content is not hostage to a single vendor's offline capabilities. It builds in two-way synchronization so a learner's progress survives the transition between offline and online, and between one location and the next.

This is the equivalent of the hospital's power grid: hard won, well designed, and absolutely necessary. The problem is not that this infrastructure exists. The problem is what it leaves unexamined.

OFDL says nothing about where its content comes from. It does not distinguish between Open Educational Resources, which are freely available and openly licensed for reuse and adaptation, and proprietary content, which is controlled by publishers, corporations, or certifying bodies that can restrict, revoke, or price access at will. It treats content as a payload to be compressed, resized, cached, and synchronized.

Educational content, however, is not merely something to be delivered. It often must be transformed. In many of the low-resource and crisis-affected communities OFDL is designed to serve, learners speak languages that are largely invisible to commercial educational publishers. A lesson available only in English or French may technically exist on the server, yet remain educationally inaccessible to the children sitting in front of it.

For many kinds of infrastructure, agnosticism would be reasonable. A shipping container does not need an opinion about what is inside it. A road does not care whether the truck is hauling grain or luxury goods. OFDL is different because it does not claim to be neutral about outcomes. Its own stated principles, including a holistic learning experience that guarantees equal opportunity for engagement, localized digital literacy built for a learner's specific environment, and continuity across the disruptions that make low-resource and crisis-affected communities vulnerable, are explicit commitments to educational equity. Whether those commitments are realized depends overwhelmingly on the content learners receive.

A brilliantly engineered offline server delivering content licensed to a single grant cycle or content that disappears the moment funding ends has not achieved educational continuity. Neither has a server stocked with proprietary materials that cannot legally be translated, revised, or culturally adapted for the communities it serves. By contrast, a server distributing Open Educational Resources (OER) provides content that can persist independently of any publisher's business model while empowering local educators to translate it, revise it, remix it, and improve it without seeking permission from distant rights holders.

In multilingual classrooms, refugee settlements, Indigenous communities, and rural schools, translation is not a convenience. It is the difference between participation and exclusion. A lesson that cannot be translated is, for many learners, no lesson at all. A science lesson translated into a child's first language does more than improve comprehension. It communicates that the learner's language and culture belong in the classroom. The ability to localize educational materials is not an implementation detail. It is one of the principal mechanisms through which educational equity is achieved. Translation transforms access into understanding, and understanding into learning. Openly licensed educational resources make that transformation legally and practically possible.

A framework that meticulously specifies its technical architecture yet remains silent about whether educators possess the legal freedom to translate and adapt the very materials they deliver has mistaken the visibility of its engineering for the completeness of its mission.

Infrastructure and educational content are not separate concerns that happen to occupy the same system. The infrastructure exists because the content matters. An offline-first learning framework can only claim to advance educational equity honestly if it has something meaningful to say about both.

OFDL has solved the engineering problem of keeping learning available during disruption. But availability alone is not educational access. Without a commitment to openly licensed, adaptable, and translatable content, the framework cannot fully realize its own goal of equitable learning. Infrastructure makes learning possible. Open, adaptable, and translatable content makes learning accessible. A framework that aspires to educational continuity cannot remain silent about content.

Sticky Fingers and Literacy Tax Dollars?

This is the time of year when Minnesota state legislators have town halls and fund raisers. At one recent event, I asked a long term state legislator if they knew how much of the money allocated by the READ Act that they had just bragged about went into the coffers of private equity companies. The response included a mumble about unintended consequences. 

State literacy initiatives are presented as investments in children. Increasingly, they have also become intentional, consequential investments for private equity (PE.) In both Mississippi and Minnesota, a small group of investment firms now owns many of the products that schools are required, or strongly encouraged to purchase. The same firms profit from screening students, training teachers, and renewing those contracts year after year. The same firms have their fingers in the legislative money pots of a lot of states.

Curriculum Associates (i-Ready), Renaissance Learning (FastBridge and Star), Edmentum (Exact Path), NWEA (MAP Growth), and Lexia's LETRS teacher-training program represent the most widely used literacy products in these states. Behind these familiar education brands are a handful of private equity firms including Veritas Capital, Blackstone, Francisco Partners, Permira, Berkshire Partners, Hellman & Friedman, and Vistria Group.  These money makers have consolidated ownership across the literacy marketplace.

This matters because these are not one-time purchases. Student screeners require annual licenses. Teacher certifications require ongoing training and renewal. Every school year, public education dollars flow back through the same recurring-revenue business model. A note here about the use of screens in school: identifying literacy needs increasingly routes through licensed software platforms, sometimes student-facing computer programs, sometimes tablet apps that teachers use to record one-on-one assessments, but in either case, the data and scoring run through proprietary, licensed vendor software rather than locally-built and locally owned and controlled programs. I guess cuz Minnesotans aren't smart enough to do that kind of thing..

A growing share of the infrastructure surrounding public literacy policy is controlled by financial firms whose legal obligation is to maximize returns for investors. When the same investment groups own the assessments that identify literacy needs and the professional development used to address them, policymakers and taxpayers should ask whether sufficient safeguards exist to ensure that educational decisions are driven by evidence rather than by market concentration.

The issue is not whether these products are effective, although that's still a very open question. The issue is accountability. When state policy effectively channels recurring public dollars into a marketplace dominated by a small number of financial firms, citizens deserve transparency about who owns these products, how much public money is spent on them each year, how those contracts are awarded and renewed, whether meaningful competition still exists, and whether the programs are actually working and not being massaged as has been widely reported about Mississippi’s numbers.

 Generating dependable returns for investment portfolios should not be the object of legislators. 

The Fear-Driven QA Trap

 When institutions or governments respond to AI-generated OER content with centralized quality gatekeeping, they're operating from a scarcity and control mindset, the same mindset that OER is designed to move away from. The logic goes: "AI tools are producing low-quality materials at scale, so we need stronger filters before anything reaches learners." That feels responsible, but it smuggles in some problematic assumptions.

The radical promise of OER isn't just free resources; it's freedom to adapt. The 5Rs framework (Retain, Reuse, Revise, Remix, Redistribute) places the teacher at the center as an active agent, not a passive consumer of vetted content. A teacher who finds a resource that is 80% right for their classroom isn't supposed to wait for a quality-assured version; they're supposed to fix the 20%.This means imperfection is not the enemy of OER; it's assumed. The whole architecture of OER anticipates that materials will need local adaptation. When quality assurance becomes the primary response to AI-generated OER content, several things happen that cut against OER's grain:

 It re-centralizes authority. A review board or quality committee becomes the arbiter of what's usable, recreating the gatekeeping dynamic of traditional publishing. That’s comforting to bureaucrats but not useful for teaching and learning.

It signals distrust of teachers. The implicit message is that educators can't judge whether a resource suits their students, that they need experts to pre-approve it for them.

It slows the ecology down. OER's strength is its velocity and adaptability. Heavy QA processes introduce bottlenecks that favor static, "finished" resources over living, iterable ones.

It mistakes polish for fitness. A highly polished resource that doesn't fit a specific classroom context is less useful than a rough one that a teacher can quickly reshape. QA processes typically optimize for the former.

Rather than asking "is this resource good enough?" before release, the OER-consistent question is "does this resource come with enough transparency for a teacher to assess and adapt it?" That shifts the work from pre-publication gatekeeping to:

Providing Clear metadata about how, when, and with what tools a resource was made;

Creating in Editable formats like Moodle courses so adaptation and localized assessments are actually possible, not just theoretically permitted. These formats need to be able to be used equally in environments with the least resources as well as classrooms with the latest technology.

Providing a Platform for Community Annotation so teachers can flag issues and share improvements contextually;

Build Educator Capacity that helps teachers develop the critical eye to evaluate AI-generated content themselves.

Using fear of low-quality AI content as a reason to reduce teacher agency is contradictory to what gives OER its beauty. Teachers know their students and are always going to be a better quality filter for their specific context than any generalized review process. Fear-based QA trades that distributed, context-sensitive intelligence for a centralized standard that fits every classroom approximately and none just right.

The better bet is trusting the OER ecosystem's own immune system which includes: teachers adapting; communities annotating; and bad resources simply not getting used or getting improved rather than building walls at the gate. Spend money on teachers in classrooms not filter creators. 

The Missing Infrastructure for OER in K-12.

 

Open Educational Resources (OER) have transformed how we think about curriculum high-quality, openly licensed materials are now available for nearly every subject and grade level. Yet despite billions invested in OER development and curation, adoption remains frustratingly slow in K-12 schools. The reason isn't content quality or teacher awareness. It's infrastructure.

The Iowa Wake-Up Call

Iowa's $17 million federal OER investment, the largest K-12 OER curriculum initiative ever, should have been a triumph. Instead, it became a cautionary tale. The state created Canvas courses featuring premium OER from Illustrative Mathematics, OpenSciEd, and Great Minds. Then, Iowa's Department of Education refused to share the materials despite their Creative Commons licenses, effectively locking publicly funded open resources inside Canvas's proprietary platform. Even Iowa's own digital learning experts were denied access.

Iowa's bungle revealed a fundamental flaw in how we're thinking about OER implementation. We've focused on creating and finding open content while ignoring the platforms that deliver, assess, and manage learning. The result? "OER-washing", open content trapped in closed systems.

Why OER Commons Isn't Enough

Many assume OER Commons solves the platform problem, but it addresses a different challenge entirely. OER Commons excels as a discovery tool; think of it as a library catalog helping teachers find quality resources. But teachers don't just need to find materials; they need to deliver courses, administer assessments, track student progress, manage grade books, and integrate with their district's student information systems.

OER Commons provides none of these capabilities. There's no student login, no assignment submission, no grading tools, no analytics dashboards. When teachers find great resources on OER Commons, they still must download them and upload them to Canvas, Google Classroom, or Schoology, returning to proprietary platforms. The discovery problem has been solved; the delivery problem remains.

The $450 Million Opportunity (and Threat)

Here's what makes this urgent: middle school math OER alone, if locked into proprietary platforms and sold nationally, represents approximately $450 million in market value. Venture capital and private equity have noticed. In recent years, for-profit companies have been positioning "open" content within closed platforms where they can charge for access, analytics, and integration features.

Districts think they're adopting free materials, but they're actually buying into expensive platform dependencies. The $XX per student they spend on LMS licenses quickly exceeds what they would have paid for traditional textbooks—except now they've lost the ability to easily share, adapt, or control their curriculum, which is the promise of OER.

The Multi-State Moodle Solution

What if states pooled resources to build shared, truly open infrastructure? A multi-state Moodle consortium could provide:

• Shared hosting and technical infrastructure that no single state could afford alone

• Common integrations with student information systems, built once and used everywhere

• Standardized professional development and implementation support

• Turnkey OER course packages that teachers can immediately use and adapt

• Genuine community control—no vendor lock-in, no surprise price increases, no surveillance capitalism

The economics are compelling. Instead of thousands of districts each negotiating separate LMS contracts, states could share costs while maintaining local control. Member fees would run far below current commercial licensing costs, and savings could be redirected to teacher training and OER customization.

Why Hewlett Foundation Is Uniquely Positioned

The biggest barrier to a unified open-source strategy isn't technical; it's coordination. States don't spontaneously collaborate at this scale. They need a trusted convener with OER credibility and the resources to fund startup costs.

The Hewlett Foundation could be that convener. With long-standing relationships across state agencies, OER networks, and education organizations, Hewlett has the credibility to bring states together. An initial $5-8 million investment over three years could establish infrastructure serving five pilot states, creating a blueprint that dozens more could follow.

This wouldn't compete with OER Commons; it would complete it, creating the delivery layer that makes Hewlett's discovery investment fully valuable. Teachers could search OER Commons, then import materials directly into their district's Moodle instance for actual teaching, assessment, and learning management.

Global Impact

The implications extend far beyond U.S. borders. A proven, well-supported Moodle model would give countries in Africa, Asia, and Latin America access to world-class learning infrastructure at a fraction of proprietary costs. Instead of each nation paying foreign vendors for platform access, they could run their own instances, customize for local contexts, and contribute improvements back to a shared commons.

Smaller language communities could finally create culturally responsive content without prohibitive platform costs. A Somali teacher in Minneapolis and a Somali teacher in Mogadishu could share materials seamlessly, exactly what OER was supposed to enable.

The Choice Ahead

Open Educational Resources stand at a crossroads. One path leads to vendor capture, where "open" content becomes raw material for proprietary profit. The other leads to genuine educational infrastructure that is community controlled, globally accessible, and truly open.

The content is ready. The need is clear. What's missing is the infrastructure and the convener to make it real. With the right leadership, we could finally fulfill OER's promise: accessible, adaptable, equitable education at scale.

The question is whether we'll build the infrastructure OER needs, or watch it get captured by the platforms it was supposed to replace.

UDL and OER: The False Promise of Universality

 I've been following the work of two educators I deeply respect and admire for almost 20 years. These are the kind of leaders every school district should have. But it's time they started complying with UNESCO's 2019 Recommendation on OER.

Their new compilation of opinions, experiences, PowerPoint presentations, and excerpts from noted educators begins with the "All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means without written permission from the editors, a specific chapter author and/or the publisher" warning.

I don't believe either educator really wants to enforce that ownership statement. They want to share ideas freely and build a community of professional teachers who exchange experiences and innovations. Instead, they're trapped by education's tired notion that ideas need a restrictive copyright to be taken seriously—a notion that serves the legacy publishing industry but fails students and teachers, especially those with the fewest resources. We must solve first for those with the least.

The Universal Design Paradox

Universal Design for Learning's claim to "universality" fundamentally fails because it ignores UNESCO's 2019 OER Recommendation. This blind spot undermines UDL's own goals and reflects systemic problems with how we distribute educational resources.

UNESCO recognizes that true educational universality demands more than accessible design—it requires resources that are freely available, legally reusable, and adaptable across all contexts without economic, technological, or legal barriers. The 2019 Recommendation calls for educational resources that can be "accessed, used, adapted, and redistributed by others with no or limited restrictions."

Yet UDL typically operates within proprietary ecosystems that contradict these principles. When UDL implementations rely on expensive proprietary learning management systems, subscription software, or copyrighted materials that cannot be modified or shared, they create a fundamental contradiction: resources designed to be universally accessible become universally inaccessible to those who cannot afford them.

Breaking the Proprietary Trap

Proprietary educational publishing undermines teacher expertise by creating closed, static systems. The "all rights reserved" model forces teachers to purchase ideas rather than share resources. Teachers cannot legally copy, adapt, or distribute content, isolating educators and stifling innovation that could make learning more accessible.

This creates a competitive environment where teachers become consumers and 'experts' gain status through commercial metrics rather than educational effectiveness. It prevents the localized adaptations and collaborative improvements that diverse learning communities require.

Open licenses transform this dynamic entirely. Teachers can revise, remix, and redistribute materials, becoming genuine partners in creating universally accessible education. An openly licensed UDL resource becomes a starting point, not a finished product. A teacher can adapt an urban gardening curriculum for desert climates while maintaining UDL's multiple means of representation, engagement, and expression—recognizing both the original author's and adapting teacher's expertise in a resource that can be shared and improved by others.

Building True Universality

OER promotes collaborative communities that align with UDL's values. When teachers improve open resources, they contribute to collective knowledge, fostering continuous improvement that makes education more accessible. This shifts the focus from "experts" to dynamic networks of professional educators addressing localized learning needs.

The adoption of OER represents fundamental trust in teachers as professionals capable of sound curriculum decisions. Until UDL integrates OER principles—making accessibility inseparable from openness, affordability, and global adaptability—its claims of universality remain hollow promises that reinforce educational colonialism.

True universal design for learning cannot exist within closed, proprietary systems that prevent the collaboration and adaptation diverse learning communities require.