Since the 1960s, a long tradition of social scientists has investigated the construction of technology and its relationship to human experience. In particular, social constructivists have questioned the alleged deterministic nature and technological neutrality, rejecting the idea that it is created in sterile spaces and immune to values. These theories, daughters of the Frankfurt School, were born in response to determinism (reductionist theory that presupposes that technology determines the structural and cultural development of a given society) and to the promoters of technological neutrality, who consider technology as a means of per itself devoid of value connotations and never right or wrong per se. A classic example is that of weapons: the defenders of neutrality did not consider them something originally connoted, but only a posteriori, based on their use.
In 1999, constructivists like Donald Mackenzie and Judy Wajcman invited us to place ourselves outside our social knowledge and understandings: the only way to define a neutral technology is through a thought experiment similar to the Rawlsian veil of ignorance. Only if it has never been used before, or if the purpose it will have to fulfill is unknown, then it is possible to judge a technological artifact as neutral.
In a famous 1980 article entitled Do Artifacts Have Politics?, the political scientist Langdon Winner proposed some ways in which politics can be imprinted in a technological artifact. The designer’s intentions, the culture of the society in which it emerges, the practical needs (and therefore the operation) of the technology itself are some of these. Winner takes the example of Long Island: according to some interpretations, overpasses were built in the 1920s and 1930s designed to prevent the passage of buses, so that only cars (and people who could afford them) could access the spaces of leisure on the island. The example is used to describe an indirect exclusive practice from the enjoyment of an asset through its own technical design, about which we tend to question little. This leads to the question to what extent the intentions of a project can remain deliberately hidden and neutral.
I am personally convinced that every technology reproduces a certain degree of power and has a political dimension. Starting from the methods expressed by Winner, some distinctions must obviously be made: we cannot compare a killer drone to one smartwatch, but not even i different uses that can be made of the same technology. One of the examples that I consider most suitable and current in this context is that of facial recognition in its various real-time or predictive applications (automatic recognition of gender, emotions, physical states).
The political dimension of biometric technology is evidently linked to its purposes of identification, surveillance and control. At this point many cities around the world, including Italy, have shown that they want to equip their cameras with this type of software to increase the security level of citizens. From the studies carried out in recent years it is the difficulty of measuring the scientific effectiveness of these systems emerged, understood as the ability to satisfy the reason why they were introduced. In most cases, these are not transparent processes, and the people potentially involved, such as citizens and residents, are not questioned. What is the precise purpose of these projects? Increase perceived security through increasingly precise video surveillance, or accurately identify people? In the second case, the example of the city of Detroit should have alarmed: according to the police chief, the facial recognition software they were equipped with could not identify people correctly 96% of the time.
Although it is one of the most invasive technologies (in a probably unprecedented way) and the protagonist of many news cases that tell of its problematic impacts, it seems very difficult to question the uses of biometric recognition. Those who criticize certain purposes and uses of a technology are increasingly accused of pessimism, Luddism, involution. The trend (as I see it) is to consider any innovation as positive regardless. It is not a provocation: we do not see scientific discussions regarding the experimentation and effectiveness of an artifact before its release on the market and in society, yet we consider each technology as something to be accepted a priori for its innovative scope. This, according to the social constructivists, happened due to the assimilation of technology to the exclusive economic and productive (and not social) progress by transforming it into something unavoidable (as for determinists), uncontrollable and unstoppable. This has slowly led us to a lack of responsibility towards its results and to the loss of critical sense, so much so that today consider as extremist the positions of those who ask that some uses of facial recognition be banned.
This is the case of the European campaign #ReclaimYourFace, which is collecting the signatures of European citizens to ask the Commission to ban facial recognition systems in public spaces. Organizations supporting the campaign (more than 50 across Europe) believe that “the use of mass biometric surveillance in member states has led to violations of European data protection law and restricted the fundamental rights of individuals, including their privacy, freedom of speech, the right to protest and not to be discriminated against “.
In support of the positions carried out by the Reclaim Your Face coalition, several institutions have recently intervened. Last month the European Parliament has spoken out against the use artificial intelligence by police and judicial authorities and against facial recognition databases, like those used by Clearview AI. The resolution is not legally binding, but provides important indications on the positioning of European parties with respect to these issues, in view of the discussion of the AI Act. Even in Italy something is moving: in Parliament it is a bill is being examined to introduce a moratorium on the use of facial recognition in public places.
According to the Vice-President of the European Parliament, Marcel Kolaja, the ban on the use of facial recognition in public spaces is an important step in the fight against mass surveillance. Wojciech Wiewiórowski, European Data Protection Supervisor (Edps), he stated in an interview with Politico that European society is “not ready” for facial recognition technologies that monitor people in public. According to the Edps, which supports an absolute ban together with other European legislators, remote biometric identification “would transform society, transform our citizens, transform the places we live in, into places where we are perennially recognizable”. Those who are pushing for the adoption of these technologies argue that they are needed to fight crime, despite the fact that the data available on its effectiveness are few and many studies tell us about them potential to exclude some people from accessing certain services.
Even Facebook has taken a stand, stating that it will erase the facial biometric data of more than a billion people and block automatic recognition of faces in photos. Meta’s Head of AI wrote in un tweet that “each new technology has costs and benefits. We believe that how to balance these in the case of facial recognition is a debate that should take place in the open ”. It takes an effort to understand that in reality the news does not introduce anything new: DeepFace’s algorithm has already been trained on that data for more than 10 years, and Meta has already made it clear that the limitation will not extend to the products of the metaverse.
In my opinion, the debate (what little there is) should revolve around specific questions: is there the possibility of introducing a new technology, but do we really need it? At what point is the state of the art? We attend software daily that instead of innovating they reproduce obsolete and despicable social dynamics. Do we really need it? The most common mistake is not to distinguish the different purposes (and therefore the reasons for the construction) of an artifact: if, as Herbert Marcuse also repeated, technology cannot be isolated from its intended use, it is very important to recognize that the same algorithm can be used for very different social purposes, such as identifying the face and unlocking the smartphone, but also by the police to identify migrants at the borders. Then, however, the intentions behind the design must be explicit and transparent, which means that a technology designed for a particular purpose must be scrutinized before it is intended otherwise.
Critical studies of the technology are not particularly progressing because we are losing the ability to question it, disprove it, and even reject it: it seems that somehow technological rationality is able to erode critical rationality. Surely the instrumental and cultural dimensions of technology cannot be separated, but there are two levels to distinguish between: the one that concerns the rules and political limits that we decide to introduce, and the world of innovation and business, which should aspire to research and evaluate impacts. What will happen until the AI Act comes into force introducing the sandbox, that is the only useful approach to guarantee innovation and risk minimization?
Criticism, by definition, is the judging art that allows us to evaluate and interpret, not to accept any hypothesis without first asking ourselves about its content and its origin. It serves to break down knowledge and research in order to solve them. This does not mean going against progress, but simply trying to adapt it to everyone’s needs. It does not mean being against technology, but rather being able to appreciate its results even more when they bring benefits. In technological criticism I see an attitude that is not only desirable, but obligatory. A task that is up to every society to renew itself, an individual and collective mobilization for a more careful evaluation of the meaning and scope of technological development.
.