Last week Microsoft confirmed, in response to reporting by TechCrunch and others, that it had handed BitLocker recovery keys for three laptops to the FBI following a valid court order. The underlying case was a fraud investigation in Guam. The laptops were encrypted with BitLocker — the full-disk encryption built into Windows, which many institutions and individuals rely on as their primary protection against unauthorised data access.

The mechanism is simple and was not widely known. When you set up a modern Windows device and sign in with a Microsoft account, BitLocker automatically uploads your recovery key to Microsoft’s cloud. No prominent notification. No opt-in. The key sits there, associated with your account, accessible to Microsoft. When a US court issues a lawful order, Microsoft complies. Redmond confirmed this is policy, not an exception.

Bruce Schneier’s response was characteristically direct: “The lesson here is that if you have access to keys, eventually law enforcement is going to come.” Jennifer Granick at the ACLU called remote key storage in this configuration “quite dangerous,” particularly given that the same mechanism is available to any government that can issue a Microsoft-compatible legal order — not only the US Department of Justice.

That last point is the one European institutions should be reading carefully.

Why This Is a European Problem

The CLOUD Act — the US Clarifying Lawful Overseas Use of Data Act, passed in 2018 — allows US law enforcement to compel US-based companies to produce data held on servers anywhere in the world. If your university stores its BitLocker recovery keys in a Microsoft account, and Microsoft is a US company, the geographic location of the servers those keys sit on does not limit a US court’s reach. The keys are in Virginia, legally, wherever the hardware is.

This is not speculation. It is the explicit structure of US digital law. The European Court of Justice has repeatedly ruled that certain US surveillance frameworks are incompatible with GDPR — the invalidation of Privacy Shield in Schrems II (2020) being the most prominent example. But court rulings about data transfer frameworks do not automatically change the operational reality for an institution whose laptops are running Windows with default settings.

European universities hold exactly the kinds of data that make this a real rather than a theoretical concern:

• Research data: medical studies, clinical trials, interviews with human subjects, social science datasets — all subject to strict ethical and legal protections
• Student records: academic performance, personal circumstances, disciplinary proceedings
• HR data: employment contracts, salary records, health information, union activity — particularly sensitive under German and EU labour law
• Correspondence and draft documents: research in progress, grant applications, peer review material

If the disk holding any of this is encrypted with BitLocker, and the recovery key has been uploaded to a Microsoft account by default, the encryption provides less protection than it appears to. The key is accessible to a foreign state with a court order. That state is not party to GDPR.

The Structural Problem

The BitLocker story is one instance of a larger pattern. It is not that Microsoft behaved unusually or maliciously — it complied with a lawful order in its home jurisdiction, as it is legally required to do. The problem is structural: when an institution depends on a closed-source, US-headquartered platform for its critical infrastructure, the institution has delegated control over its own data to an entity whose legal obligations lie elsewhere.

This applies beyond encryption. It applies to email (Exchange Online, Outlook), document storage (SharePoint, OneDrive), communication (Teams), identity management (Azure Active Directory), and any service that runs through a Microsoft account or Azure tenant. For each of these: the data is subject to Microsoft’s terms, and Microsoft is subject to US law.

The same argument applies, with different specifics, to Google Workspace and any other US-headquartered platform. The issue is not that these companies are bad actors. It is that their legal accountability and the legal accountability of European public institutions point in incompatible directions, and the institutions mostly have not noticed.

What Sovereign Software Looks Like

The alternative is not paranoia and air-gapped servers. It is a coherent strategy for institutional digital infrastructure that is based on software the institution controls.

In Germany, this conversation has a name and a project. OpenDesk — developed under the aegis of the federal and state governments — is a stack of open-source tools (Nextcloud, Collabora Online, Matrix/ Element, Jitsi, Keycloak, Open-Xchange) assembled into an integrated workspace alternative to Microsoft 365. The Souveräner Arbeitsplatz (sovereign workspace) concept behind it is exactly what the BitLocker story illustrates: if the software is open, the keys stay in your institution, and no foreign court can reach them via a warrant served on a US company.

Several German states and federal agencies have been piloting OpenDesk. The city of Munich’s earlier experiment with Linux (LiMux) and its eventual rollback to Windows is the cautionary tale here — not because open source failed, but because the transition was not supported seriously enough over time, and the incumbent vendor’s lobbying was. The BitLocker story is a reminder of what is at stake in that political negotiation.

The FSFE’s “Public Money? Public Code!” campaign has articulated the principle cleanly: software developed with public funding should be released as open-source software. The argument is not only about freedom as an abstract value. It is about the practical consequence of being locked into a proprietary platform: your institution loses the ability to audit what the software does, to modify it to meet your requirements, to host it where your data protection law applies, and to switch providers without losing access to your own data.

What I Do, and Why

I work at a publicly funded institution. The software I build for institutional contexts — campus infrastructure, workforce management, archival systems, alert systems — is public.

Not because I am ideologically committed to open source as a movement, but because the alternative is incoherent. If I build tooling for a university with public funds and keep it closed, I have produced a private asset with public money, duplicated by every institution that builds the same thing independently, inspectable by nobody, and ultimately dependent on my continued willingness to maintain it or hand it over. None of those outcomes serve the institutions I am building for.

Here is what that looks like in practice:

zammad-ticket-archiver — automated archival of Zammad support tickets as cryptographically signed PDFs, with RFC 3161 timestamps for non-repudiation. Built for institutions that need legally defensible audit trails of their helpdesk operations. The signing infrastructure is self-hosted; no external party holds the keys.

alarm-broker — a silent panic alarm broker for campus facilities. Receives emergency triggers from hardware devices (Yealink keys), distributes notifications via Zammad, SMS, and Signal, with acknowledgment tracking and escalation scheduling. Runs locally, logs to self-hosted PostgreSQL; no external dependency for the alarm path.

campus-app-kit — a React Native / Expo starter for university mobile applications, with a pluggable Node.js backend designed for institutional data sources (room booking, events, schedules). The architecture separates institution-specific connectors (which institutions keep private) from the shared foundation (which is public). Any university can take it and build on it without starting from scratch.

cueq — an integrated workforce management system for German universities under TV-L (the collective agreement for public sector employees in the German states). Handles time recording, shift planning, absence management, payroll export, and GDPR-compliant audit trails. Built around NestJS and Next.js, with a PostgreSQL backend and Honeywell terminal integration. The HR data stays on the institution’s own infrastructure.

These are all boring. They are not research contributions; they are plumbing. But plumbing is what holds institutions together, and the question of who controls the plumbing — and under whose legal jurisdiction — is exactly the question the BitLocker story makes visible.

The Principle

Public money, public code. If an institution funded by public money develops software for its own operations, that software should be released under an open licence, inspectable, forkable, and deployable by any institution with the same needs.

The corollary: institutions funded by public money should prefer software that is itself openly licensed, auditable, and deployable on infrastructure the institution controls. Not as a blanket ban on proprietary tools where they are genuinely the best option, but as a starting presumption that shifts the burden of justification.

The BitLocker story is not a story about Microsoft doing something wrong. It is a story about the logical consequence of a procurement decision that was made without asking “and what happens when a US court sends a subpoena?” That question was available in 2018 when the CLOUD Act passed, in 2020 when Schrems II was decided, and before both. It is still available now, for every institution that has not yet asked it.

The FSFE “Public Money? Public Code!” campaign is at publiccode.eu. The OpenDesk project is at opendesk.de. The original TechCrunch reporting on the BitLocker handover is at techcrunch.com.

This story is not supported by the evidence.

The relationship between socioeconomic background and educational outcomes is one of the most replicated findings in social science. PISA data from Germany consistently show one of the steepest socioeconomic gradients in the OECD — the correlation between parental education and student performance is higher here than in most comparable countries. This is not a recent finding. It has been stable for decades. The system produces it reliably, which means the system is, in some meaningful sense, designed to produce it — even if no individual actor intended that design.

Understanding why requires a different vocabulary than the one most educational institutions use about themselves.

Bourdieu’s Three Capitals

Pierre Bourdieu spent much of his career developing an account of how social inequality reproduces itself through culture and education. The core concept is capital — but not only the economic kind.

Bourdieu (1986) distinguishes three forms:

Economic capital is material resources: money, assets, time purchased through money. This is the most visible form of advantage. Wealthier families can pay for tutoring, for better-resourced schools, for the unpaid internships that build CVs, for the years of postgraduate study that increasingly function as the entrance requirement for professional careers.

Cultural capital is more subtle. It includes dispositions, skills, and knowledge that are valued by educational institutions and professional fields — but valued in a way that tends to favour those who acquired them at home, in childhood, before formal education began. The ease with which a student navigates a seminar. The familiarity with the tacit conventions of academic writing. The sense that the university is, broadly, a place made for people like you. These are not things that are explicitly taught; they are things that are transmitted, Bourdieu argues, through families whose own cultural capital aligns with what the institution expects.

Social capital is networks: the web of relationships that provide information, referrals, opportunities, and vouching. Who you know, in the flattest possible terms.

All three reinforce each other. Economic capital can be converted into cultural capital through education and into social capital through exclusive networks. Cultural capital eases access to prestigious institutions, which build social capital. The system is not static, but it has a strong gravitational pull toward reproduction.

Bourdieu and Passeron (1977) developed this into a theory of education as reproduction: the function of educational institutions is not primarily to transmit knowledge but to legitimate the transmission of social position from one generation to the next. The process is misrecognised — by students, teachers, and institutions — as meritocracy. This misrecognition is essential to the function. If it were transparent, it would lose its legitimising power.

The Hidden Curriculum

Philip Jackson (1968) coined the term hidden curriculum for everything that schools teach that is not in the official syllabus. How to sit still. How to wait your turn. How to speak to authority. How to navigate institutions, read implicit expectations, manage bureaucracies. How to understand that your job is to demonstrate competence within a form that someone else has set.

For students whose home culture aligns with the institutional culture, the hidden curriculum is invisible. They already know it; it requires no effort; it is simply how things are. For students whose home culture diverges, it is a second curriculum that must be decoded while simultaneously managing the official one.

Lareau (2003) documented this in careful ethnographic detail. Middle-class families engage in what she calls concerted cultivation — a mode of child-rearing that practises precisely the dispositions valued by educational institutions: articulate self-advocacy with adults, a sense of entitlement to ask questions and seek explanations, activities structured around developing discrete skills. Working-class and poor families, in her study, more often practised accomplishment of natural growth — providing security, affection, and freedom without the institutional structuring. Neither is better parenting. But one of them is what the school expects.

The child who arrives at school already knowing how to talk to teachers, how to present themselves, how to advocate for their own needs, has a significant advantage that is invisible in the transcript. It does not appear as “privilege”; it appears as “ability” or “maturity”. The institutional category does the misrecognising work.

Privilege as Invisible to Those Who Have It

Peggy McIntosh (1989) wrote what became one of the most cited — and most contested — essays in education: “White Privilege: Unpacking the Invisible Knapsack”. Her core observation is structural: privilege is the absence of disadvantage, and absences are invisible to those who live inside them. You do not notice the ease with which you move through a system that was designed for people like you, any more than you notice breathing.

This is not an accusation. It is a description of a structural feature with consequences for self-understanding.

My background is in physics; I now work in universities, having grown up in a household with books and educated parents and the background assumption that higher education was something that people like us did. I was not aware of most of this as an advantage while it was happening, because it did not feel like an advantage — it felt like normal. The awareness came later, with effort, and it remains incomplete.

The invisible entrance fee is what you have already paid, in cultural capital, before you walk through the door. The institution does not ask about it explicitly. It simply rewards those who have it and attributes the reward to merit.

What This Means for Accessibility

The previous post in this series argued that full accessibility — Barrierefreiheit — is structurally impossible in a society organised as ours is; that the honest goal is Barrierearmut, the ongoing reduction of barriers. The connection to privilege is direct.

Barriers to education are not only physical. They include the cultural distance between the home environment and the institutional culture. They include not knowing that office hours exist and are meant for you, not just for students with problems. They include the inability to identify as “the kind of person who does a PhD” because you have never met anyone who did one. They include the exhaustion of navigating a system that requires you to translate yourself at every step, while your better-resourced peers spend that cognitive energy on the actual work.

None of these barriers appear on an accessibility audit. They are not visible from inside the institution looking out. They require actively listening to people whose experience differs from the institutional default, and then being willing to revise the default rather than add an exception.

The PISA gradient in Germany is a measurement of accumulated, unreduced barriers. It is not a measurement of the distribution of talent or effort. The system is producing the outcome; the students are receiving the label.

The Meritocracy Problem

Meritocracy is an appealing concept and a damaging ideology when taken seriously. The appeal: rewards should go to those who earn them, and earning should depend on effort and ability rather than inherited position. This is genuinely better than aristocracy.

The problem: in a society with steep inequality in the distribution of cultural, economic, and social capital, “merit” is not a neutral measurement. It is a measurement of the match between a person’s accumulated resources and the demands of the institution. Calling that match “merit” names the outcome without naming the process that produced it.

Michael Young, who invented the word “meritocracy” in 1958, intended it as a satire. His book The Rise of the Meritocracy depicted a dystopia in which the illusion of fairness made inequality more entrenched, not less, because it stripped the legitimacy from those who were left behind. If outcomes are fair, then failure is your fault. The ideology provides the institutional absolution; the individuals bear the moral weight of structural disadvantage.

This is precisely the dynamic that Bourdieu’s theory of misrecognition describes. The student from a poorly resourced background who does not reach the outcomes of their better-resourced peer is seen — by themselves, by teachers, by the institution — as less talented or less motivated, rather than as navigating a steeper gradient with fewer tools.

What Institutions Can Actually Do

The structural critique is not an argument for fatalism. Institutions can do things that matter.

They can make the hidden curriculum visible — explicitly teaching what is usually assumed. That means orientation programmes that actually explain institutional culture, not just procedures. It means academic writing support that is not remedial but normative. It means mentoring that connects first-generation students with people who understand the landscape.

They can audit their practices for whose default they assume. The timed closed-book exam was designed for a particular set of conditions; asking what it actually measures, and whether there are better instruments, is not lowering standards — it is interrogating what the standard is measuring.

They can diversify their faculty and staff, not as a cosmetic gesture but as a structural change in whose tacit knowledge is embedded in the institution. If the people who design the curriculum all navigated it from the same starting position, the curriculum will encode that starting position as normal.

They can name the entrance fee. Acknowledging that outcomes correlate with background, that this is a systemic feature and not a distribution of merit, is the first step toward taking institutional responsibility for the gradient rather than attributing it to the students.

None of this resolves the structural problem. The structural problem requires political change at scales well beyond any individual institution. But institutions are not passive. They can reduce the barriers they control, while being honest about the ones they do not.

A Personal Note

I sit in institutional positions that this analysis would identify as advantaged. I teach in a university. I benefited from the gradient in ways I cannot fully account for. The point of naming this is not guilt; it is responsibility. Being advantaged by a system you did not design does not make you complicit in its worst outcomes — but it does make you responsible for using whatever institutional leverage you have to make the system less exclusive.

The connection to accessibility is this: both inaccessibility and privilege are about whose defaults are built into the system and who is required to adapt to defaults they did not set. Reducing barriers and interrogating privilege are the same project, approached from different angles.

Neither is completable. Both are necessary.

References

• Bourdieu, P. (1986). The forms of capital. In J.G. Richardson (Ed.), Handbook of Theory and Research for the Sociology of Education (pp. 241–258). Greenwood Press.
• Bourdieu, P. & Passeron, J.C. (1977). Reproduction in Education, Society and Culture. Sage. (Original French edition 1970.)
• Jackson, P.W. (1968). Life in Classrooms. Holt, Rinehart and Winston.
• Lareau, A. (2003). Unequal Childhoods: Class, Race, and Family Life. University of California Press.
• McIntosh, P. (1989). White privilege: Unpacking the invisible knapsack. Peace and Freedom, July/August, 10–12.
• OECD (2023). PISA 2022 Results (Volume I): The State of Learning and Equity in Education. OECD Publishing.
• Young, M. (1958). The Rise of the Meritocracy. Thames and Hudson.

The short answer is that YouTube does not work fine, for reasons that matter specifically to universities. The longer answer is what this post is about.

What Is Wrong with YouTube

YouTube is the world’s dominant video platform. It is free to use, globally available, handles any file size, transcodes automatically, and comes with an audience of two billion logged-in users. For individual creators who want reach, it is genuinely hard to beat.

For universities it fails in at least three important ways.

Data protection. When you upload a lecture or a concert recording to YouTube, the content, the metadata, and the viewing behaviour of your students go to Google’s servers — which are predominantly in the United States. Under the GDPR, transferring personal data to third countries requires either an adequacy decision, standard contractual clauses with additional safeguards, or explicit informed consent. After the Schrems II ruling (Court of Justice of the EU, 2020), the adequacy of US-based data transfers became legally contested in a way that makes institutional YouTube use genuinely difficult for European universities. Using it for anything with identifiable students — which includes most teaching content — is a compliance problem.

Platform logic. YouTube is designed to maximise watch time. Its recommendation algorithm is not neutral. It will recommend whatever comes after your lecture, and what comes after your lecture is not under your control. For educational content — especially sensitive material, or content that should remain in a defined pedagogical context — this is a real problem. The platform is not indifferent to what is hosted on it; it shapes how it is consumed.

Institutional fragility. YouTube is free until it isn’t. Platform terms change; monetisation policies change; content is demonetised or removed based on automated systems with imperfect appeal mechanisms. Building institutional infrastructure on a free commercial service is a bet that the commercial incentives of that service will remain aligned with your needs. That bet has a poor historical record.

None of this means that individual instructors should never use YouTube. It means that universities should not make YouTube their default institutional solution.

educast.nrw: A Cooperative Model

educast.nrw is a project of the Digitale Hochschule NRW — a cooperative of North Rhine-Westphalian universities building shared digital infrastructure. The concept is a state-wide video service, run by universities for universities, for the recording, processing, management, and distribution of video content in teaching and research. The platforms calls itself “Hochschul-YouTube”, which is both accurate and slightly underselling what makes it different.

The HfMT Köln participates as a user institution, which is where I come in as the IT contact for setting it up on our end.

The technical foundation is Opencast, an open-source video management system developed by a consortium of universities. This matters: the software is auditable, the development direction is set by the institutions that use it rather than by advertising revenue, and the infrastructure runs on German servers that are explicitly GDPR-compliant. Licenses on uploaded content are freely choosable — CC-BY-SA is an option, which means the university’s teaching materials can be open access if that is what the instructor wants.

What It Can Do

The feature set covers the actual use cases of a university, not the use cases of a content creator trying to build a following.

Recording. The Opencast Studio browser app records in three modes: screen only, camera only, or screen and camera simultaneously as a synchronised multi-stream. That last option — Presentation and Presenter as separate streams, played back with the viewer switching focus between them, or as picture-in-picture — is the format that works for a lecture. You get the slides and the speaker in the same video, but the viewer can choose which to focus on. That flexibility is not something you get from a simple screen recording uploaded to YouTube.

Multi-perspective video. For a music university this is the feature that changes things. A concert or a masterclass is not well-served by a single camera angle. The platform supports simultaneous recording from two camera perspectives — a wide shot and a detail shot, say, or a front view and a hands view for a piano performance. The viewer can switch between them in playback, or the institution can set a default presentation. This is infrastructure that makes the teaching use of concert recordings actually feasible, not just technically possible.

Formats. Video up to 4K with adaptive bitrate streaming (the player adjusts automatically to the viewer’s connection), audio up to broadcast quality (48kHz/16-bit, exceeding CD’s 44.1kHz), with FLAC at up to 96kHz/24-bit in development. No file size limit. These specifications matter for music. A piano recording compressed to whatever YouTube decides to do with it is not the same as an uncompromised audio stream. The difference is audible.

ILIAS integration. This is the practical hinge. Video that lives in a separate platform is video that students may or may not find. Video embedded directly in the ILIAS course page, in the learning module, at the point in the curriculum where it is relevant — that is video that is part of the course rather than adjacent to it. The integration between educast.nrw and ILIAS is direct: upload to the video platform, embed in ILIAS on pages, in learning modules, or as standalone video objects, all from within ILIAS.

Access rights. The granularity here is what distinguishes it from any public platform. Each video can be set to: public (anyone on the internet), institution-wide (anyone logged in at the university), course-wide (only enrolled students in a specific course), individual (specific named people), or private (only the uploader). A graduation concert might be public. A practice session for student feedback might be course-only. A recording made for an individual student’s reflection might be shared only with that student and their teacher. These are all normal use cases in a music university; they all require different settings; the platform handles all of them.

Use Cases in a Music University

The general university use case — lecture recording, video tutorial, self-study module — applies at HfMT as much as anywhere. But a music and dance university has some specific ones.

Concert recordings. HfMT Köln runs performances at both its Cologne and Wuppertal sites. Recording these and making them available to students, faculty, and selectively to the public used to mean someone had to manage files, find hosting, deal with YouTube’s automated copyright detection flagging student performances of copyrighted repertoire, and explain to students why their graduation concert had been muted by an algorithm. The controlled platform makes all of this manageable.

Stage presence as a reflective tool. Watching yourself perform is a standard part of performance training. It is uncomfortable, useful, and until recently required either dedicated recording equipment or the ad-hoc use of a phone propped against something. A proper recording infrastructure with controlled access — the student sees the video, their teacher sees the video, nobody else does — changes the pedagogical viability of this approach. The barrier to actually using video feedback in practice teaching drops substantially.

Theory and practice. This is the institutional argument I made in the presentation and stand by: video infrastructure that works for a lecture also works for a concert. The same system that stores the introduction to music theory also stores the masterclass by a visiting artist. This is not incidental — it is the point of a shared infrastructure. You do not need to choose a platform for academic content and a different one for performance content. The platform works for both.

The Argument Behind the Argument

There is a broader principle at work here that extends beyond video platforms. Public universities are funded by public money. The infrastructure they build with that money — software, platforms, data, content — should be under their control, governed by their values, and ideally available to other public institutions. The commercial platform model inverts this: you get free hosting in exchange for your data, your students’ attention, and your institutional dependence.

educast.nrw is an example of what the alternative looks like in practice: a cooperative of public institutions building shared infrastructure on open-source software, governed collectively, with data on European servers under European law. It is not perfect — the setup overhead is real, the user experience does not match YouTube’s, and the feature roadmap (automatic subtitling, H5P support, livestreaming, annotation tools) is still catching up to what commercial platforms have had for years. But the model is right.

The question of who owns the video infrastructure of a university is the same question as who owns its email, its learning management system, its student data. The answer should be: the university, operating under law, answerable to its students and to the public that funds it.

The slides from the Tag der Lehre 2022 presentation are available on request. For educast.nrw setup at HfMT Köln, contact the IT department.

Links

• educast.nrw
• Opencast
• Opencast Studio
• Tobira Videoportal
• opencast-ilias plugin (GitHub)

Changelog

• 2025-08-18: Corrected the capitalisation of “OpenCast” to “Opencast” throughout (matching the project’s official spelling on opencast.org).