Conjure Utopia

The word DeCoTech has been floating around more and more in recent times: it stands for “Democratic Control of Technology”, and it’s the idea that digital technology should be conceived, designed, developed, maintained, and monitored with the participation of the people who are going to use it, who are going to be impacted by it, or who have an explicit democratic mandate from the communities and individuals involved.

In this article, I will summarize the concept concisely since I see it used often behind closed doors, or referenced under this or other names in pamphlets, blog posts, books, and papers that never take the time to lay down exactly what we are talking about. This includes a recent strategy paper I published, which made me realize both the galvanizing power of this concept and the need to systematize what it means.

Currently, most new digital technology in the West is produced without democratic oversight. A few centralized investors direct its creation, and a network of big or small private actors organized in business ecosystems execute the vision. State actors intervene by either investing money to reinforce the private sector in alignment with their agenda, supplying their technologies, or regulating such technologies after they have been released on the market. State actors with a liberal regime often act within this space with limited accountability to the public. The consequence is that policymaking is determined by the conflict between specialized civil society and tech lobbying groups.

It’s a bit more complicated than that, but we are not here to talk about the status quo. We are to talk about what we want to build.

Democratic Control of Technology is an umbrella term for a series of radical changes within the current power structure to shift control of technology towards workers, users, communities, and old or new public institutions. It is not a single demand, nor a blueprint, but a general term to identify a winning condition for political actors. Since Democracy is not a binary condition, but a fluctuating property of a system, a spectrum along which we can evaluate the fairness of a process, DeCoTech can be achieved through a multiplicity of paths and result in many different new political entities and institutions.

DeCoTech is not a new invention: the implicit or explicit idea of having democratic oversight over digital technology has existed for a long time. DeCoTech, though, is useful to bring together viable answers to the problems of the current moment. It is a catalyst to aggregate political forces towards a shared direction.

Now, all this talk about power structures, actors, and systems is very abstract. Leaving DeCoTech as a general, pluralistic term, let’s look at some existing phenomena that express a degree of DeCoTech.

Let’s start with environments in which the users and the developers come together to design a certain technology or its deployment. Software developed through participatory design, data cooperatives, platform cooperatives, or any situation in which related communities are treated as stakeholders with equal power to the developers is an expression of local DeCoTech.

Another form of existing DeCoTech is tech workers’ resistance against corporate decisions on technology. Examples include No Tech for Apartheid, resisting their employers’ collaboration with the genocidal military forces. In a non-democratic setting like private corporations, the expression of democracy passes through workers resisting the authoritarian power of the employers, forcing them into complicity with the development of harmful technology. While the people directly impacted by the technology are not involved in the process (also because such technologies are designed to kill them, and I imagine the governance meetings would feel quite tense), the tech workers stand as a replacement through a desire for solidarity and justice.

Some degree of DeCoTech is also obviously expressed by state regulations around technology, even though they have limited and delayed impact. This is primarily due to institutional capture by tech lobbying and limited political will to oppose private interests in favor of the popular interest.

Now, what would an increased level of Democratic Control over Technology look like? There is an endless array of proposals, which I won’t even try to categorize. Some want to put the State in charge of planning technological development but under new institutions. Some want to put workers’ power in the workplace at the center. Some others want decentralization of technology at the community level, with every neighborhood vibecoding their own hyper-custom stack. Some proposals are an overarching economic overhaul; some are concerned with single, narrow aspects.

An example I like to bring up when I speak about DeCoTech is the way, as a society, decided it was good to test all medical drugs before putting them on the market. In the 19th century, the proposition of extended tests for all commercial drugs was unthinkable. It took several incidents with thousands of victims to lead to the formation of dedicated institutions to validate what was going to land on the market. The whole industry reshaped, and not necessarily for the better: the costs of testing led to a huge concentration of capital in what we now call Big Pharma. Despite this and other shortcomings, we can all agree that we prefer to have the guarantee that the medicines we buy at the pharmacy have had a degree of testing.

Why can’t we do it for commercial software? Well, we are already doing it for a small portion of it. There’s auditing for privacy reasons, or for critical software that, if it malfunctioned, would kill a lot of people, like the one that runs planes or cars. These auditing systems have been under attack for a long time by corporate interests, but they show that where there’s a will, there’s a way. CEOs, if presented with this proposition, would kick and cry endlessly and make up claims that it’s technically unfeasible. It’s easy to understand why they don’t like this kind of scrutiny: code won’t be an industrial secret anymore, and it will be under continuous public scrutiny. No more shady practices, dark patterns, reckless deployments, secret surveillance of users or workers, and a permanent end to the “move fast and break things” ethos.

Private companies needing to ask for permission, rather than forgiveness, is but a small building block, a first step towards DeCoTech.

Verifying and validating the impact of all the software put in production can be a daunting task, and most likely depends on the State having an incentive to protect Democratic power, which, these days, doesn’t seem to be the case. Big Tech has captured many State institutions in the USA and Europe, embedded itself into the military, and fundamentally sabotaged the electoral process, Janonism notwithstanding.

Let’s approach DeCoTech from a different perspective. Tech Workers within tech companies are the potential frontline of DeCoTech for the very nature of their position in the productive process: they observe directly how every decision of investors and managers is translated into requests to alter software. A degree of DeCoTech happens when tech workers, collectively, decide to contest “how” a technology is developed. A much higher degree of DeCoTech happens when they contest the “what”, such as the previous example with No Tech For Apartheid.

A high degree of DeCoTech will be achieved when tech workers have, in most companies, the power to consistently decide what technology is developed and for what purpose, in coordination with the adopters of such technologies. This will require entirely new institutions to be designed in a way that the development process itself is conditioned and constrained by the consensus and approval of all parties impacted. This is not a trivial effort, and I don’t want to offer a simplistic and underwhelming solution here, such as a deliberative assembly or a wishy-washy cooperativemaxxing.

One last branch of DeCoTech I see being discussed is in the spaces of economic planning: appropriating the logistics technologies that, for example, Amazon or Walmart use to predict and plan a good portion of the global economy, and using them in combination with the plenty of economic planning models and simulators to move away from the unstable, unreliable, and self-destructive economic system we live under. CyberSyn in Chile, despite its enormous limitations, was definitely a good example of DeCoTech.

I want to conclude by highlighting how a lot of people have the feeling DeCoTech will become a central point of discussion in the immediate future, at least in Europe. While I’m writing, the stocks of many tech companies are in freefall. I don’t know if this is the AI bubble bursting or not yet, but it’s clear that the current trajectory is not sustainable. Tech Bros are to blame. More generally, the neoliberal order is shattered, and whatever Trump is going for doesn’t seem built to have a lasting impact if not within the USA, then at least in the rest of the world. Politicians don’t have good answers and are entrenching themselves either in a gruesome defense of the dying order or in a frontal, irrational assault on the remnants of the establishment.

Soon or later, there will be a strong demand for a new way of doing things, and we need to be ready with a compelling vision to heal society from the madness we are going through.

Given the harm and destruction Big Tech has brought in the last few decades through surveillance, militarization, pollution, and the capture or destruction of public institutions, DeCoTech is not only a political project, but a necessary condition to insulate society from the damage digital technology is capable of, and to turn this potential in service of the fight against the enormous challenges that climate collapse and other ecological threats pose to civilization.