After two years spent diversifying its fields of action, La Quadrature du Net is now tackling a new front: the fight against the onslaught of artificial intelligence (AI) in all areas of society. To help keep the critique of authoritarian and ecocidal digital technologies alive, La Quadrature more than ever needs your support in 2025.
For several years, along with other collectives in France and Europe, we have been documenting the very concrete consequences of the growing adoption of artificial intelligence across various sectors: through the Technopolice and France Contrôle campaigns, or even more recently with investigations into the environmental impact of data centers powering the exponential growth in storage and computing power.
A triple capitalist accumulation
In recent months, following the sudden hype of generative artificial intelligence and products such as ChatGPT, and under the aegis of big business and complicit states, we have witnessed a further acceleration of the computerization process. This acceleration is the direct consequence of everything that is already problematic about the dominant trajectory of digital technologies. First of all, the huge accumulation of data over many years by major tech multinationals such as Google, Microsoft, Meta and Amazon, who monitor us to better predict our behavior, and who are now capable of indexing gigantic corpora of text, sound and images by appropriating the common good that is the Web.
Collecting, storing and processing all this data requires a prodigious accumulation of resources. This is reflected first and foremost in capital: the rise of tech, boosted by surveillance capitalism, has attracted the favor of financial markets and benefited from accommodating public policies. Thanks to this capital, these companies can finance near-exponential growth in the data storage and computing capacity needed to drive and run their AI models, by investing in graphics chips (GPUs), undersea cables and data centers. These components and infrastructures in turn require immense quantities of rare earths and metals, water and electricity.
With this triple accumulation – of data, capital and resources – in mind, it’s easy to see why AI is the product of everything that’s already a problem in the digital economy, and how it adds to the bill. However, the marketing (and media) myth of artificial intelligence deliberately conceals the issues at stake and the intrinsic limits of these systems, including the most efficient ones (biases, hallucinations, waste of resources necessary for their operation).
Exploitation squared
The political and media hype surrounding AI overlooks the concrete effects of these systems. Far from solving humanity’s current problems thanks to the allegedly superior rationality that would emerge from its calculations, AI in its concrete uses amplifies all existing injustices. In the economic sphere, it translates into the massive and brutal exploitation of hundreds of thousands of “data workers” tasked with refining models and validating their results. Downstream, in the organizations in which these systems are deployed, it induces a new seizure of power by managers over workers in order to increase corporate profitability.
Of course, there are some relatively privileged workers in the service sector, or even in the “creative classes”, who now see this as an opportunity to “save time” in a society sickened by the race for productivity. It’s a new “dictatorship of convenience”: on an individual level, everything encourages us to be complicit in this logic of collective dispossession. But rather than freeing workers, it’s a safe bet that the automation of work induced by the increasing use of AI will, in fact, contribute to further accelerating work rates. As was the case with previous waves of computerization, AI is also likely to be accompanied by a dispossession of knowledge and a de-skilling of the professions it affects, while contributing to wage cuts, worsening working conditions and massive destruction of skilled jobs – at the same time worsening the precariousness of huge swathes of the population.
In the public sector too, AI is accentuating the automation and austerity already affecting public services, with deleterious consequences for social cohesion and inequality. National education, where the “pedagogical” AI of a startup founded by a former Microsoft executive has been tested since September 2024 and without any prior evaluation, appears to be a particularly sensitive area where these developments are already at work.
Busting the myth
To support the myth of “artificial intelligence” and minimize its dangers, an emblematic example is systematically highlighted: it would be able to interpret medical images better than a human eye, and detect cancers faster and earlier than a doctor. It could even read analysis results to recommend the best treatment, thanks to an encyclopedic memory of existing cases and their specificity. For the time being, however, these tools are still in the development stage, and will only supplement doctors’ knowledge, whether in reading images or helping with treatment.
Whatever their actual effectiveness, “medical” use cases act in the AI mythology as an isolated heroic moment that in reality conceals an altogether different political agenda. This strategy of mystification can also be found in other fields. For over twenty years, governments have been using the need to combat child pornography and terrorism to justify the surveillance of communications. In the mythology of police algorithmic video-surveillance (AVS), it is the example of the little girl lost in the city, and found in a few minutes thanks to cameras and facial recognition, that is systematically used to convince people of the merits of total video surveillance of our streets.
We need to put aside the screen of the virtuous example and show the unavowable uses that AI apostles would rather hide, and document the ensuring pernicious reduction in rights and freedoms. We have to realize that, as an industrial paradigm, AI increases the harm and violence of contemporary capitalism tenfold. It aggravates the forms of exploitation already enslaving many of us. AI multiplies state violence, as illustrated by the increasing place given to these devices within military apparatuses, as in Gaza where the Israeli army uses it to accelerate the designation of targets for its bombardments.
Tracing alternatives
Instead of fighting AI and its evils, public policies in France and Europe today seem essentially designed to consolidate the hegemony of the tech industry. This is notably the case with the AI Act or “AI regulation”, despite being touted as a bulwark against the dangers of “excesses”, when in fact it seeks to deregulate a booming market. In this age of Startup Nation and absurd praise for innovation, most leaders see AI as a lifeline, a Holy Grail capable of saving Europe from economic collapse.
Again and again, it’s the argument of geopolitical competition that has been mobilized to silence critics: whether in the report of the French Government Committee dedicated to generative AI or in that of Mario Draghi, it’s all about flooding multinationals and start-ups with cash money. The goal is to make sure Europe stays in the race led by the United States and China. This all comes down to treating evil with evil, by repeating the mistakes we’ve been making for over fifteen years: more and more “magic money” for tech, while public services and the commons are subjected to austerity. It’s a choice to roll back protections for rights and freedoms, so that AI can proliferate throughout society.
These policies are absurd because everything suggests that Europe’s industrial lag in AI cannot be made up, and that this race is therefore lost in advance. Above all, these policies are dangerous insofar as, far from being the life-saving technology they often claim to be, AI actually accelerates ecological disaster, amplifies discrimination and increases many forms of domination. The current paradigm not only locks us into an unsustainable headlong rush, but also prevents us from inventing an emancipatory political trajectory in line with planetary limits.
Although AI is presented as inescapable, we don’t want to resign ourselves to it. Faced with the soft consensus that reinforces a devastating capitalist system, we want to help organize resistance and sketch out alternatives. To do so in 2025, we need your support. So if you can help us, go to don.laquadrature.net!