AI interview: Thomas Dekeyser, researcher and film director
On the politics of ‘techno-refusal’, and the lessons that can be learned from a clandestine group of French IT workers who spent the early 1980s sabotaging technological infrastructure
April 1980, Toulouse. As the offices of multinational technology firms Philips Data Systems and CII-Honeywell-Bull sit burning, French authorities frantically search for the culprits of the suspected arson and bombing attacks.
While there was some confusion in the immediate wake of the destruction over who was responsible, with militant left-wing group Action Directe initially claiming responsibility, a communique issued by a clandestine group of IT workers cleared up the matter by detailing the exact contents of the Philips Data Systems director’s office desk.
Known as the Committee for Liquidation or Subversion of Computers (or Clodo, a play on French slang for homeless), the same communique also outlined the group’s motivations for carrying out the attack.
“As you will have suspected, we are IT workers, and are therefore well-placed to understand the current and future dangers of IT and telematics. The computer is the preferred tool of the dominant. It is used to exploit, to file, to control, to repress,” they wrote.
“This is what we are attacking, and will continue to attack. Our sabotage is only a more spectacular version of those attacks performed daily by us or by others … We don’t want to be locked in the ghetto of programs and organisational platforms. Fighting against all dominations is our only goal.”
While Clodo is not widely remembered in the same way as other radical groups from the time – which tended to be much more violent and less focused on IT as a tool of oppression – a 2022 documentary titled Machine in Flames has prompted significant interest in the group, as well as their tactics and motivations.
Speaking with Computer Weekly, Machine in Flames co-director Thomas Dekeyser – who is also the author of an upcoming book on the politics of “techno-refusal” – said he became interested in Clodo because they represented a markedly different approach to the stifling bureaucracy of today’s digital politics.
“We think of technological politics as perhaps pushing for specific legal regulations of particularly harmful elements of certain technologies. We may think of union organising within large tech companies, we think of prosecuting Zuckerberg for privacy violations, and so on,” he says.
“For the longest time, I’ve felt like those things are important elements of any political or social struggle, but at the same time business does continue as usual, despite those set of actions by well-meaning individuals, collectives and institutions, we still end up with different versions of the same problems.”
Dekeyser’s interest in Clodo, therefore, stems from not seeing their activities as the ultimate solution to society’s problems, or something to be particularly glorified, but from their “much stronger sense of refusal” around technology as a starting point for political action.
He adds that given the current social and political climate, there is something “alluring” about the immediacy and anonymity of Clodo’s actions, which prompted him to “explore forms of digital politics that do not come back every single time to the traditional political subject that we think of – whether it’s the worker, whether it’s the law, whatever it may be”.
Using Clodo as an initial jumping-off point, Dekeyser says his upcoming book is therefore an attempt to historically trace the rich range of forms resistance to technology has taken.
“Sometimes it does come back to the state, sometimes it goes back to the worker, but the intention is to go back into history and figure out what other modes of resistance to technology there have been in the past. I am interested in asking: how might these resonate with the present in ways that expand how we currently think of treating the most harmful elements of something like AI?”
Techno-refusal and being human
Dekeyser says a major motivation for writing his book is that people refusing technology – whether that be the Luddites, Clodo or any other radical formation – are “all too often reduced to the figure of the primitivist, the romantic, or the person who wants to go back in time, and it’s seen as a kind of anti-modernist position to take”.
At the most basic level, he argues that the history of techno-refusal is heterogeneous and diverse, with different historical actors having completely different motivations for opposing particular technologies at certain points in time.
“The subjects that resist are so diverse, and so are their motivations for refusal,” he says, adding that ‘technophobe’ or ‘Luddite’ have long been used as pejorative insults for those who oppose the use and control of technology by narrow capitalist interests. “I want to push against these terms and what they imply.”
Despite the manifold motivations behind techno-refusal, Dekeyser says a common feature binding the different forms of resistance is philosophical: “The refusal of technology and technological advancements are tied up with a struggle for what it means to be human.
The subjects that resist [tech] are so diverse, and so are their motivations for refusal
Thomas Dekeyser
“Technology can reinforce and undermine your ideas of the ‘ideal human’. The shared theme that pulls together the various cases in the book is the idea that technology always does something to how we understand what we mean by ‘the human’. My argument is that this partially explains why technology has drawn, and continues to draw, such incredible amounts and intensities of attention and debate, with people willing to fight for or against it with sometimes incredibly strong conviction.”
For Dekeyser then, the history of technology is necessarily the history of its refusal. From the Ancient Greek inventor Archimedes – who Dekeyser says can be described as the first “machine breaker” due to his tendency to destroy his own inventions – to the early mercantilist states of Europe backing their guild members’ acts of sabotage against new labour devices, the social-technical nature of technology means it has always been a terrain of political struggle.
“Mercantilist economics is much more protectionist than what we now think of as capitalism, so that means that they were very protective of the guilds and crafts,” he says, adding that early mercantilist support for such actions evaporated with the emergence of Western capitalism in the early 19th century, as states, technology and economics become much more tightly intertwined than ever before.
As this paradigm shift was underway, a systemic and organised approach to machine breaking also began in 1811 with the Luddites, a collective of weavers and textile workers who carried out their workplace sabotage in response to the unilateral imposition of new technologies (mechanised looms and knitting frames) by a new and growing class of industrialists.
Changes in ruling class composition as a result of burgeoning industrialisation therefore “meant that states were no longer supportive of workers breaking machines, but were in fact prosecuting now those machine breakers, which is what you see with the Luddites who faced mass hangings”.
Dekeyser adds that there was another tightening of the relationship between technology, capital and the state with the emergence of cybernetics during the Second World War, which was described by its founder Norbert Wiener as a science of communication and control.
“The basic idea is that one can control different kinds of systems – social systems, natural systems, economic systems and so on – by heightening the level of information that’s shared between elements,” he says. “There was a lot of writing in the 70s and 80s around how society could be managed and controlled, including how revolutions can be stopped, by just making sure you have enough information. Information gathering and information sharing is at the heart of the cybernetic logic.”
Cancelling the future
Dekeyser says that while the use of the term cybernetic has fallen out of vogue, the diffusion of data gathering technologies, sensors and mechanisms into everyday life means that overall approach is now baked into modern forms of governance and technology, particularly artificial intelligence (AI).
“[The emergence of the cybernetic logic is] a crucial historical moment. That’s why in the book I focus on Clodo as an attempt at countering that cybernetic ideal of the human and society,” he says, adding that this logic is particularly harmful because of how it destroys the future, “in the sense that it aims to anticipate and predict everything through data gathering”.
With the proliferation of AI and other automated algorithmic technologies through the global economy, Dekeyser sees it as a natural extension in the logic of cybernetic computation, in that it’s diminishing our very ability to conceive of and create radically different futures.
“Because AI can only ever generate outputs based on existing data, it has a strong tendency to repeat existing ways of thinking about and representing things, making it much harder to envision totally new futures,” he says.
Because AI can only ever generate outputs based on existing data, it has a strong tendency to repeat existing ways of thinking
Thomas Dekeyser
“AI is the perfect example of something that displaces our idea of the human, because AI more than any other technology or set of technologies we’ve encountered is very much a problem of many people feeling like they are losing agency in the face of infrastructures capable of doing things previously taken to be uniquely human.”
Further highlighting the deterministic nature of AI technologies – which take data from the past, with all its biases and assumptions, and project them into the future – Dekeyser says the goal of AI’s “ultra-cybernetic” logic is to neutralise and stabilise unpredictable outcomes or events for the benefit of capital and the state.
“That, for example, totally depletes the possibility of massive protests or a proper threat to the state. The thing with policing is even if it’s not an immediate outspoken aim, it’s the gathering of data that makes sure that whenever things start getting out of hand, it will be known, prevented and policed,” he says.
“I do think of these technologies as a cancellation of the future because it’s set up to predict and predetermine what the future will look like, even if this may not always work.”
Direct action and Clodo’s politics of techno-refusal
Throughout their various communications, Clodo highlighted the destructive nature of the post-war cybernetic logic, specifically rallying against the role of computing in the prosecution of imperialist wars (mainly Vietnam), conducting police surveillance, and maintaining the ideological and economic domination of the capitalist class, explaining in one communique that “the progress of technology is not the same as the progress of humanity”.
They also took aim at the dehumanising role IT can play in the workplace, which they argued massively contributed to the increasing drudgery of nine-to-five work.
Dekeyser says Clodo’s actions are significant because they challenged the cybernetic view of the human as a system to be controlled or managed while also highlighting the physical materiality of computer technology, and thus its vulnerability – far from being some ethereal thing beyond a human’s physical grasp, it is made up of real objects such as servers, fibre optic cables and microchips.
“What’s unique about certain forms of direct action is that it’s leading by example, but also signifying its ideas through action,” he says. “It specifically shows that these digital technologies, which seem so omnipresent and yet unable to be grasped, are literally cables in the ground and have addresses.”
While Clodo would go quiet for two years after the April 1980 arsons, they carried out a series of further attacks on multinational computer firms and government data processing sites throughout 1983. This included setting fire to the offices of US computer manufacturer Sperry for its complicity in then-president Ronald Reagan’s invasion of Grenada, and bombing a data processing plant in the Toulouse suburb of Colomiers.
In another communique issued in the wake of the Colomiers attack, the group celebrated the destruction of files and databases belonging to the local authority and police: “The brain drain continues! Last night, at more than 6,000 meters/second, a fraction of the state’s memories dissipated into air at Colomiers.”
While French authorities denounced Clodo at the time for endangering human life, the group’s acts of sabotage never harmed another human being – something they made clear they were avoiding in various public statements.
They also highlighted the interplay between technology, capital and the state in these statements, which explained their perspective on IT and computers as a tool of oppression.
In a 1983, for example, one anonymous Clodo member “interviewed” themselves to defend their actions: “The truth about computerisation should be revealed from time to time. It should be said that a computer is just a bunch of metal that serves only to do what one wants it to do, that in our world it’s just one more tool – a particularly powerful one – that’s at the service of the dominators.
“We are essentially attacking what these tools lead to: files, surveillance by means of badges and cards, instrument of profit maximisation for the bosses and of accelerated pauperisation for those who are rejected.”
In that same “interview”, the Clodo member noted that when “faced with the tools of those in power, dominated people have always used sabotage or subversion”.
Dekeyser adds that part of the political appeal of direct action is the immediacy of the activity and the empowerment it gives to those carrying out the actions, while consciousness-raising politics tend to be built around the rational exchange of ideas.
“If you’re just doing something out of a commitment to the rational exchange of ideas, it can be really hard. Petitioning, for example, can be the most soul-destroying thing you could possibly do. Again, it has its value in certain times and in certain spaces but, my god, it’s the perfect route to burn out – there’s no excitement to it,” he says.
“You’re just like, ‘Okay, this is me is walking around in the rain, just trying to convince people of these ideas’.”
He further notes that, to this day, no Clodo members have ever been identified or broken ranks to unmask themselves, which is part of the group’s allure: “They were not interested in personal gain, or getting voted in as MEPs. It was not about the self, purely about the politics.”
‘We’re not in Kansas anymore’
However, Dekeyser notes that at the time Clodo was active, computers tended to be the preserve of massive corporations and state authorities such as the military or police, which made attacking them much more straightforward.
“If you are to attack computers now, let’s say you go and cut fibre optic cables as people have done in the past few years in France and Germany, you’re possibly ruining thousands of ordinary people’s days,” he says. “There’s been a saturation of these technologies to the point that the object of critique and object of attack is much harder to define. Therefore, it’s much harder to get general support for it.
“Computation today feels more benign because we’re all using it on a day-to-day basis, which was not the case in the early 80s when you didn’t have personal computers yet. At that point, it was really easy for Clodo to go, ‘Computers equals government authority, corporate control, policing counter revolution, and that’s it’. And so militancy becomes much, much harder [today] because people go, ‘Yes, it’s terrible, but it’s also really important to how I live my life, how I work and everything’.”
Despite the abundance and interconnectedness of technology in modern life making it much harder to define the object of attack and critique, Dekeyser notes that at our current historical juncture, most technology is controlled by just a handful of corporations, meaning people are still finding the potential for different kinds of directed militancy.
“You see it with people organising, specifically against Amazon, particularly against Google. In Berlin, you had the campaign called ‘F*ck off Google’, which was just about preventing a Google campus from being built there,” he says. “It’s a kind of militant action – it involves occupation, it involves blocking buses from going into the campus. So, these forms are definitely emerging.”
Even with these changes, Dekeyser adds that practices of technological sabotage similar to Clodo’s are continuing into the present day, which also work to highlight the materiality of computer technologies.
One example is the so-called Vulkangruppe, or Volcano Group, which has targeted fibre optic cables, electricity pylons and tech company-owned vehicles, mainly around Berlin. In March 2024, for example, the group claimed responsibility for a further attack on electricity pylons that shut down the Tesla Gigafactory in Grünheide, explaining their motivations in a statement published online.
However, to Dekeyser’s point about the diffusion of modern technology, the resulting power outage also affected 6,500 homes, raising the question of whether groups like these will ever be able to achieve a certain level of popular support, which in turn raises further questions about the political aims being strived for.
“Should your aim be mass consciousness raising, or should it be something different?” he says. “Clodo, for example, were not interested in reaching a big audience, they were just like, ‘This sucks, we’ll burn it down’ … gaining popular support to lead to change is a fair strategy, but it’s obviously not the only one. That’s what these groups are showing us.”
Prefigurative technology
Despite the qualms many would have with the actions of collectives such as Clodo or the Volcano Group, and the clear inconvenience technological sabotage causes in today’s ultra-interconnected world, Dekeyser says they are still making a crucial political point about the prefigurative nature of technology.
While “prefigurative politics” refers to the idea that means cannot be separated from ends, and that any action taken to effect change should therefore be in line with the envisioned goals, and not reproduce existing social structures or problems, Dekeyser says this also extends to technologies, which are bound and prefigured by the logics underpinning their development and use.
The idea that just putting harmful technologies in the hands of workers or a ‘good state’ is likely to be insufficient in resolving the problem
Thomas Dekeyser
“The idea that just putting harmful technologies in the hands of workers or a ‘good state’ is likely to be insufficient in resolving the problem. These [capitalist and military] logics sit at the heart of the technology, you can’t just carve that out,” he says, noting how the French Communist Party at the time was dismissive of Clodo for “fighting against the tools of labour” on the basis that these computer technologies held emancipatory potential for workers.
Clodo, however, took the view that simply adopting existing technologies for different purposes would only replicate the same problems.
“I think this is what’s so crucial about groups like Clodo – they argue, for example, that just nationalising a technology or bringing it into public ownership, while an improvement, still leaves questions about what you are going to do with the surveillance mechanisms or military impulses that sit at the very heart of their functioning,” he says.
“So, to the question of alternative technologies, we need something that doesn’t just put something that already exists to a different use, but actually tries to start from scratch.”
Giving the example of social media platforms, Dekeyser adds that while it may be the case that data gathering is inherently baked into the technology, and that it may never be possible to have meaningful social interaction in the digital sphere without leaving some sort of traces, the starting point of an alternative would be to begin with a rejection of the data gathering impulse.
He adds that while it’s not about finding the purest position – “because that will never be possible” – it’s about ensuring technologies are not produced by the same influences: “It’s the classic thing about prefigurative politics, where you’re trying to undo something at the same time as you’re trying to build something else.”
Read more about digital politics
AI interview: Dan McQuillan, critical computing expert: Critical computing expert Dan McQuillan speaks to Computer Weekly about the top-down imposition of artificial intelligence on society, and how AI is a fundamentally political technology shaped by humanity’s most reactionary tendencies.
AI interview: Gavin Mueller, media studies professor: Author Gavin Mueller speaks to Computer Weekly about the history of resistance to automation, and why Luddite perspectives and practices are useful in pushing for more equitable digital futures.