The internet was built by the United States military. This is not a conspiracy theory — it is the matter-of-fact starting point for any honest history of the network. Understanding what the internet is requires understanding where it came from, what it was designed to do, and by whom. It also requires understanding what emerged against that grain — which turns out to be substantial, real, and ongoing.
Both things are true. The internet is a surveillance and military infrastructure. It is also a terrain on which communities have fought, and sometimes won, genuine liberation. Knowing both histories is the beginning of using the network with clear eyes.
The Pentagon Built It#
ARPANET — the direct ancestor of today’s internet — came online in 1969, funded by the Defense Advanced Research Projects Agency (DARPA). Its designers were not utopians. They were Cold War engineers solving a military problem: how do you build a communication network that survives a nuclear strike? The answer was packet switching — breaking messages into fragments that route themselves around damage. If one node is destroyed, the packets find another path.
This architecture was not designed for privacy, anonymity, democratic participation, or community control. It was designed for military resilience. The infrastructure that underlies every email, every website, every encrypted message you send today was paid for by US taxpayers and built to serve the strategic interests of the United States government.
Yasha Levine’s Surveillance Valley (2018) traces what this origin story actually means in practice: the behavioral surveillance tools, the counterinsurgency applications, the never-severed relationship between the internet’s core infrastructure and the national security apparatus. The ARPANET included surveillance tools from the beginning. The NSA’s PRISM program — revealed by Edward Snowden in 2013 — was not a sudden departure from the internet’s nature. It was its nature, grown to full expression.
What Emerged Against the Grain#
Within that military infrastructure, and in explicit opposition to it, communities built something different.
The MIT hacker culture of the 1950s and 60s developed a set of values — openness, sharing, the belief that information should flow freely and that systems should be improvable by anyone who understood them — that had nothing to do with military hierarchy. These early hackers were not criminals. They were the people who stayed up all night because they were genuinely curious about what the machine could do. (See our Hacker Culture page for this history in full.)
The Free Software Movement, launched by Richard Stallman in 1983, was an explicitly political response to the privatization of software. The GNU General Public License (GPL) was a legal weapon designed to ensure that software, and all its derivatives, would remain permanently free — not free in price but free as in freedom: the freedom to run, study, modify, and share. Stallman understood that proprietary software was not just a business model but a form of control.
Bulletin Board Systems (BBSs) in the 1980s were genuinely local community networks — dial-up systems run by hobbyists, serving their immediate neighborhoods. Before the commercial web, these were the internet’s social layer: imperfect, idiosyncratic, and actually owned by their communities.
The cypherpunks of the late 1980s and early 90s — a loose network of cryptographers and activists — understood early that surveillance was the internet’s default mode, and that encryption was the political response. Their motto: cypherpunks write code. Phil Zimmermann’s PGP (Pretty Good Privacy), released in 1991, gave ordinary people access to military-grade encryption for the first time. The US government called it a munition and tried to prosecute Zimmermann. They failed.
Indymedia, launched in 1999 during the WTO protests in Seattle, built a decentralized, volunteer-run independent media network using the internet to publish news that corporate media wouldn’t touch. At its peak, Indymedia had hundreds of local nodes worldwide — a genuine grassroots alternative information infrastructure.
The Zapatistas — the EZLN, an Indigenous revolutionary movement in Chiapas, Mexico — used the early internet to broadcast their 1994 uprising to the world and build international solidarity networks. A colonized and militarily outgunned Indigenous community had turned the US military’s communication infrastructure into a tool for their own liberation. This was not accidental. It was deliberate and strategic.
Corporate Capture#
Tim Wu’s The Master Switch (2010) documents a pattern that recurs across information technologies: a period of openness, during which creativity flourishes and power is distributed, followed by consolidation in the hands of a few who capture the technology and close it off.
The internet followed this pattern with startling speed. The open protocols of the 1990s — email, the web, IRC, Usenet — gave way to the walled gardens of the 2000s and 2010s. Facebook replaced community forums. Google replaced open search. Amazon replaced independent e-commerce. Twitter replaced Usenet. Each replacement was technically smoother and more convenient. Each was also a transfer of power from distributed community ownership to centralized corporate control.
The result is what researchers call platform capitalism: a handful of corporations own the infrastructure through which most human communication, commerce, and coordination now flows. They set the terms. They can silence, surveil, monetize, or simply disappear any community that uses their platforms. They answer to shareholders, not users.
The open internet did not disappear — it still exists, and the Collective uses it. But it became the road less traveled, requiring more knowledge and more effort to access and maintain. The default became the corporate platform.
The Surveillance Apparatus#
One of the most important and uncomfortable facts about the internet’s liberation tools is that they share infrastructure — and sometimes funding — with the surveillance apparatus they resist.
Tor, the anonymizing network used by journalists, dissidents, and privacy-conscious communities worldwide, was developed with US Navy funding and continues to receive significant funding from the US State Department. This is not a reason to stop using Tor — its code is open, its design is sound, and its actual users are overwhelmingly ordinary people seeking privacy, not government assets. But it is a reason to hold liberation tools with clear eyes rather than uncritical faith.
The surveillance infrastructure did not emerge after the fact. Levine’s research shows behavioral surveillance tools in ARPANET’s original architecture. The NSA built its mass surveillance systems into the backbone of the internet — the cables, the routers, the data centers — over decades, largely without public knowledge. When Snowden revealed PRISM in 2013, he was not describing something new. He was describing something that had been built quietly, for years, using the same infrastructure that carries everyone’s traffic.
This does not mean resistance is futile. It means resistance requires knowledge of what you are resisting.
Indigenous and Community Counter-Movements#
Some of the most significant fights over internet sovereignty have been waged by Indigenous nations.
Marisa Elena Duarte’s Network Sovereignty: Building the Internet Across Indian Country (2017) documents how tribes in the American Southwest — facing corporate telecoms that had no interest in serving rural Indigenous communities — built their own internet infrastructure. Tribal broadband networks, owned and operated by the communities they serve, are not just a connectivity solution. They are an assertion of sovereignty: the right to control the communication infrastructure on which contemporary life depends.
The Māori community’s Te Pā Tūwatawata data center (documented in the Relational Infrastructures for Sovereign Data Storage zine, IDA/Te Kāhui Raraunga, 2026) takes this further — infrastructure shaped by tikanga Māori (Māori customary values) and governed by Māori community processes, not settler colonial frameworks. The technology is not neutral. The governance makes it relational rather than extractive.
The broader Indigenous data sovereignty movement — articulated through frameworks like the CARE Principles (Collective Benefit, Authority to Control, Responsibility, Ethics) — insists that data about Indigenous peoples, lands, and knowledge must be governed by Indigenous peoples. Not stored by corporations. Not analyzed by outside researchers. Not sold to agencies. Governed by the communities whose lives it describes.
The Land Needs Data — And Land-Based People Should Own It#
Consider a tribal environmental monitor documenting eel habitat along a stretch of the Klamath for a dam removal recovery study. She knows which backwater channels the eels use to overwinter because her grandmother told her, and her grandmother knew because she had watched the river her whole life, and that watching went back further than any written record. That knowledge — hyperlocal, intergenerational, empirically verified across generations — is the most precise data in the room. But when she sits down to map it, she is using ArcGIS Online, a subscription service owned by Esri, whose servers are in Redlands, California. When the grant ends, the license expires. The maps live on a corporate server. The knowledge her grandmother’s grandmother carried, now encoded in a shapefile, is behind a paywall the tribe may or may not be able to afford next year.
This is not a hypothetical. It is the working situation of tribal environmental departments, language programs, cultural resource offices, and community organizing efforts across Indian Country and in every bioregional community doing serious land and water work. The data that matters most — salmon return counts, habitat maps, traditional use documentation, environmental baselines that underpin treaty rights claims and conservation strategies — is being generated by people with irreplaceable local knowledge and then stored in systems they do not control, analyzed with tools they do not own, managed by contractors who leave when the funding cycle ends.
The failure modes of this dependency are specific and recurring:
A GIS contractor hired to map ancestral territory does not know which boundary line matters for a treaty rights argument and which is an administrative artifact from a 1950s federal survey. They do not know that one place-name on the map is the settler approximation of a Karuk name that means something entirely different about that site’s use and significance. They produce a technically correct map that is culturally illiterate, and when the contract ends, the institutional knowledge of why the map was built the way it was leaves with them.
A database vendor does not know that certain species location data is sensitive — that disclosing where a particular plant grows could enable poaching of a culturally significant resource, or that a specific fishing location is tied to ceremonial use and governed by protocols about who may know it. They treat all records as technically equivalent, because for them they are.
A corporate software subscription creates a permanent dependency. Every year, the community must pay to access its own data. Every platform update may break workflows. The vendor’s product roadmap determines what is possible. And the data itself — the observations, the records, the maps built from years of community knowledge — lives on servers the community does not own, subject to terms of service written by lawyers in another state, accessible to law enforcement subpoena without the community’s knowledge or consent.
When outside contractors and corporate platforms hold the technical infrastructure of a community’s ecological and social work, the community has reproduced in digital form the same dependency that colonial extraction always creates: the knowledge and labor flow in, and the control flows out.
Traditional ecological knowledge is not separable from the protocols that govern it. In living Indigenous knowledge systems, knowing something comes with obligations about how that knowledge moves — who holds it, who may access it, under what circumstances it is shared, what relationships are required before it is disclosed. A database built by someone who doesn’t understand these protocols will either exclude TEK entirely (because the contractor doesn’t know how to ask) or incorporate it carelessly, stripping it of its relational context and making it available in ways the knowledge-holder never intended.
The person best positioned to build a database that properly handles Karuk traditional ecological knowledge is someone who already understands what that knowledge is, how it moves through the community, and what obligations surround it. That person may be a fisheries monitor, a language keeper, a fire practitioner, a basket weaver who also watches the birds. They can also learn to be a database administrator. A GIS analyst. A network administrator. A security practitioner. These are learnable skills — not magic, not reserved for a credentialed class of outsiders. They are tools, and tools can be taught.
The skills that ecological and social sovereignty require right now include:
- GIS and spatial data — who controls the maps of a territory controls how that territory is understood, managed, and contested in courts, agencies, and the public record. Communities with their own GIS capacity do not depend on outside agencies to describe their own land.
- Databases — for cultural knowledge, genealogy, species observation, environmental monitoring, language documentation, resource management. The difference between community-controlled infrastructure and a corporate platform is who decides who can see what, and under what conditions.
- Groupware and coordination tools — organizing across a geographically dispersed rural community requires shared tools. Those tools can be self-hosted and controlled, or they can be Google Workspace, where the terms of service mean that everything is available to federal subpoena and corporate data analysis.
- Encryption and communications security — the documentation of treaty violations, the records of a land rights campaign, the locations of culturally sensitive sites, the organizing conversations of a community resisting extraction: all of this has adversaries, and those adversaries have the same tools law enforcement and corporations have always used to surveil and disrupt movements.
- Network infrastructure — community-owned mesh networks, tribal broadband cooperatives, local servers: infrastructure that does not route through corporate systems, that stays up when commercial providers go down in a wildfire, that is governed by the community rather than a distant corporation.
None of this requires that every community member become a full-time technologist. It requires that technical skills exist inside the community — that there are people who hold them, who are rooted in the place and accountable to the people, who will still be there after the grant ends and the contractor has moved on.
This is why collective, community-based tech literacy is a dimension of sovereignty, not a nice-to-have. The goal is not to produce isolated individual experts — it is to build shared capacity that circulates through the community, held by many hands, not dependent on any single person or any single grant cycle. Knowledge shared between neighbors is harder to extract, harder to defund, and harder to dispossess than knowledge held by a contractor.
What the Collective Is For#
The Klamath Tech Collective is a response to exactly this situation, in exactly this place.
We are not neutral about technology. We think the surveillance infrastructure is real and the tools to resist it are learnable. We think communities that depend on corporate platforms and outside contractors for their data infrastructure have reproduced colonial dependency in digital form. We think the people who know this watershed — who know the salmon runs and the oak stands and the eel channels and the burning schedules — are the people who should be building and governing the data infrastructure of this bioregion, not handing that work off to someone from outside who will be gone in eighteen months.
That means our work is about more than digital self-defense for individuals, though it includes that. It is about building durable technical capacity inside the community — capacity that persists across grant cycles, that is held by people with real stakes in this place, that is informed by the knowledge and the protocols that living here creates.
We want to teach GIS to the people who are already monitoring habitat, so that the maps they build reflect what they actually know and live in community-controlled infrastructure. We want to help tribal environmental departments and language programs build databases that run on their own servers, governed by their own data policies, inaccessible to outside parties without explicit community consent. We want to teach encryption to land defenders and water protectors who cannot afford for their documentation and organizing to be intercepted. We want to help community organizations run their own groupware rather than feeding their internal communications into Google’s data infrastructure. We want to build local mesh networks — like our Solar-Powered Mutual-Aid Comms Network — that keep communities connected when the commercial infrastructure fails during fire season, and that don’t route through anyone else’s servers.
This is what it looks like when the counter-history of the internet is not just a history but a practice: people in a specific place, rooted in it, building the technical infrastructure of their own sovereignty rather than outsourcing it to the same extractive systems the history describes.
A traditionalist who knows the river and also knows how to query a database, build a secure map, and encrypt a document is harder to dispossess than one who must ask an outside contractor for every data task. A community that runs its own servers, governs its own data, and can defend its own communications is practicing a form of sovereignty that no grant cycle can take away.
There is also a role for settlers and non-Native people who hold these technical skills — not to lead this work or arrive as experts, but to share skills when asked, to show up as learners-alongside rather than contractors-in-charge, and to understand that making technical knowledge genuinely available to Indigenous and land-based communities who want it is a form of relational allyship. Projects like the Indigenous Mapping Collective have modeled what this looks like in practice: technically skilled people in genuine solidarity with communities doing their own mapping, on their own terms, transferring capacity rather than extracting data. Growing that kind of relationship — bringing skills into the community rather than extracting community knowledge into outside systems — is one of the Collective’s explicit goals.
The internet was built by the Pentagon. Cypherpunks and Zapatistas and tribal broadband cooperatives and Māori data center builders and Indymedia volunteers did extraordinary things with it anyway. The fight over who controls the infrastructure, the data, the protocols, and the knowledge of how it all works is ongoing. It belongs to us as much as anyone — and in this watershed, it is ours to take up.
Further Reading#
- Levine, Yasha. Surveillance Valley: The Secret Military History of the Internet. PublicAffairs, 2018.
- Rankin, Joy Lisi. A People’s History of Computing in the United States. Harvard University Press, 2018.
- Wu, Tim. The Master Switch: The Rise and Fall of Information Empires. Knopf, 2010.
- Coleman, Gabriella. Coding Freedom: The Ethics and Aesthetics of Hacking. Princeton University Press, 2013.
- Evans, Claire L. Broad Band: The Untold Story of the Women Who Made the Internet. Portfolio, 2018.
- Duarte, Marisa Elena. Network Sovereignty: Building the Internet Across Indian Country. University of Washington Press, 2017.
- Indigenous Data Alliance & Te Kāhui Raraunga. Relational Infrastructures for Sovereign Data Storage. IDA, 2026.
- Indigenous Mapping Collective — community-led mapping initiatives centering Indigenous peoples’ relationships to land
Also on this site: Technocolonialism · Hacker Culture · #Data Back · Animism, AI & Consciousness
Want to build these skills together? Come to a workshop or reach us at contact@klamathtech.diy .
Stay connected: