Digital resources in the Social Sciences and Humanities OpenEdition Our platforms OpenEdition Books OpenEdition Journals Hypotheses Calenda Libraries OpenEdition Freemium Follow us

ai and the Middle East, assignment complementing chapter “classification” (deadline May 24)

Controversial technology analysis. Research and write a blog post about a use of AI classification/identification technology in the region (facial recognition, etc.) and discuss the pros, cons and ethics involved.


OpenEdition schlägt Ihnen vor, diesen Beitrag wie folgt zu zitieren:
Victoria Mummelthei (21. Mai 2024). ai and the Middle East, assignment complementing chapter “classification” (deadline May 24). Keine Disziplin – No Discipline. Abgerufen am 7. März 2026 von https://doi.org/10.58079/13jqt


8 Antworten auf „ai and the Middle East, assignment complementing chapter “classification” (deadline May 24)“

  1. The “Classification” chapter was very rewarding to read, as I found it answered many thoughts and questions I had about the biases and behavior of AI and its classification methods.

    It made me think of many examples, including the AI-driven automated border control systems Turkey uses. The country’s airports, particularly Istanbul Airport, have implemented facial recognition technologies following the introduction of biometric passports, which have been in use for nearly a decade. Many people I know, including myself, have used these automated customs gates numerous times. The gates are quite similar to the EU citizen gates at Berlin Airport customs. They are exclusively for biometric Turkish passport holders and, in addition to the facial recognition setup, they also use fingerprint scanning machines.

    I prefer these gates because they significantly reduce the time spent in the queue to enter the country. This has advantages, such as reducing wait time. A Turkish passport holder already has their biometric data in the passport, so using the gates could be seen as utilizing that already-provided data. Nevertheless, this could also be problematic, as the data could easily be used against citizens. Non-electronic passport holders are still subjected to facial recognition systems at the unautomated custom gates. It seems that there is no way to avoid being exposed to AI when entering the country.

    The collection of extensive individual data can be seen as either an advantage or a disadvantage. While the system aims to improve the efficiency of border control, it can also be used for scrutiny. It may be viewed as increased security for the country, or as extreme surveillance. There are also concerns regarding privacy and the use of this data. I think it creates a reality of constant observation and analysis of people, which is a threat to personal freedom.

  2. The growth and adoption of facial biometrics and other artificial intelligence technologies in the Middle East region, specifically in UAE and Saudi Arabia has grown exponentially. This highlights the implementation of these technologies in various sectors, in addition to their impressive growth within the facial recognition market over the past few years, which has been demanded by key industries and advanced infrastructure projects.

    Among these projects that have been carried out recently, we can distinguish those related to security and surveillance, with the installation of biometric systems to improve the security of people who decide to make the pilgrimage to Mecca. And on the other hand, the implementation of facial recognition cameras in Dubai for events such as the 2020 Expo. Also, the concern about smart cities is increasing in this area. NEOM in Saudi Arabia is one of the current examples integrating AI technologies to create mega-connected communities and improve efficiency in both logistics and transportation for a more optimised future development.

    On the one hand, the advantages of analysing its use within this region is the potential to help reduce crime and thus improve public safety through such surveillance. On the other hand, one of the disadvantages framed, from my point of view, not beneficial at all, is the massive collection and storage of biometric data. This raises serious concerns about privacy and the possible misuse of the personal information obtained. Not only would privacy be affected, but excessive and constant surveillance can affect individual liberty to levels that can be detrimental to the security of oneself and the people around one’s environment. Thus, technological dependence would increase, and people and many cities would be vulnerable to technical failures or cyber-attacks, for instance.

    Therefore, the use of facial biometrics and AI in the Middle East raises important ethical questions. These systems need to avoid any kind of racial or gender discrimination, and those individuals being monitored need to be informed and consent that data collection. Lack of transparency in these cases can have a direct impact on society and cause irreversible damages. There must be a strong legal framework to regulate the use of these technologies to protect citizens rights and ensure their ethical and responsible use.

    So, while facial recognition and AI offer significant benefits in terms of security and efficiency, they also present challenges in terms of privacy, ethics and equality that must be carefully addressed to ensure fair and safe use of these technologies in the future.

  3. In addition to facial recognition and racial profiling, how AI is being used and where investment is going in the MENA region creates another kind of classification between classes and countries. In a booklet published by Microsoft in 2019, there is a detailed list of which countries are receiving AI investments in which areas. (https://info.microsoft.com/rs/157-GQE-382/images/report-SRGCM1065.pdf) Here, the report is based on the experiences of executives from companies using AI in different sectors, such as health and banking. Under the category of defining and using AI, some of the subcategories are interesting, such as biometrics, speech recognition and computer vision. Biometrics is explained as: “Analysis of human physical and emotional characteristics – also used for identification and access control”, and computer vision is: “Gives computers the ability to ‘see’ images as humans see”. However, it is not clear how AI processes this data. Nor is it clear how AI understands human emotional characteristics. When it comes to the amount of money in terms of investments, it seems that Turkey is the most invested country in terms of AI with the amount of 252 deals and $3459 million. All these investments also show that there is a hierarchical structure between MENA countries, not only in terms of AI itself, but also in terms of the money used to fund it. The report also states that “when it comes to AI, knowledge is power” and “business-minded people will drive the transformation”. It is obvious that this report is not focused on the benefit of the general public, but rather on companies and their executives. If AI is power, or at least the knowledge of AI, then these investments and the transformation that comes with AI show us one thing: the class structure in the MENA regions is changing drastically in favour of the companies and the already wealthy class that has access to this kind of knowledge. “AI leadership” is one of the most used word combinations in this document. As far as I can see, when it comes to AI technology, there is a strict categorisation between those who can use the data and the technology and the others.

  4. Crawford’s chapter of “Classification” delves into the prejudices that underlie within AI systems, ones in which undermine the very social and political benefits that have been marketed to the public. In opening, Crawford discusses the case of polygenism and its worldwide implications on “intelligence and claims to legal rights acts as a technical alibi for colonialism and slavery” (Crawford, 126). Although the study of polygenism far predates the technological surge of what contemporary AI has become, it notes critical correlations in the ways in which the classification politics produce discriminatory results based on its discriminatory patterns that are trained into the data. Crawford uses a few examples to corroborate this statement, one of those including the case of ImageNet. Classification politics through the search engine of ImageNet shows the discriminatory patterns of categorization within hierarchies of gender and sexuality. The data in which they classified sexuality was through “homosexuality as a mental disorder in the Diagnostic Manual” (ibid, 138). It shows that embedded within its data, ImageNets holds prejudices that target marginalized groups and perpetuates repressive ideologies.

    In more contemporary AI findings, Crawford alludes to facial recognition and the ways in which it is used to exploit peoples privacy and utilize its data in facial recognition. This puts minority communities at risk as racial prejudices within politics (asper the redlined communities that continue to exist in the United States) lead to greater surveillance in such neighborhoods. The final example that stuck out was the role of classification politics in social media platforms like TikTok, Instagram, Facebook, and X, and Google. On these platforms there is little known in how users are targeted and monitored. Although this ambiguity could be perceived as docile, this is not the case “as they fail to offer meaningful avenues for public contestation” (ibid, 149). Ultimately, Crawford believes that political interference must be implemented before having AI enter the public sphere, however, I find that even within the political sphere, prejudices and classification politics can still take precedence and permit these technologies to operate without restriction. I believe, user transparency and civilian contestation must be permitted in order to change the classification hierarchies that exist within AI; this of course is only a first measure of recourse, further understandings and research must be conducted in finding what successful means is necessary in alleviating such prejudices.

  5. In her book, Kate Crawford delves into the past and present of scientific classification systems. She examines how racism, masked as scientific objectivity, had manipulated science, and she applies this critique to AI technology. She questions how biased perspectives in historical practices like skull measurements reemerge in AI today. In short, AI systems build databases from collected data, but what if this data is biased?
    Although it’s been asked to discuss a use of AI classification from the ‘region’ , first thing occurred in my mind after reading the chapter was my own experiences through ads, as being an individual from the ‘region’. Kate Crawford uses the term ‘classification’ to mean physical appearance, but it is also possible to be classified through only use of a smartphone.
    Since I have been living in Germany, I like to observe how ad algorithms work. The ads I have to face always give me the impression that what I have been defined is more important than what I actually am.
    When I am on mobile phone and have to watch a video with an ad, it’s normal to be interrupted by a charity ad for building a new DITIB Mosque in southern Germany, where a hodja explains how good it is for Muslims to help other Muslims.
    Or, when I read an article on my phone, a dating app founded for German-Turks pops up, with a German and Turkish flag, featuring a woman with a hijab on one side and a blonde woman without hijab one on the other, as if they are like shoulder angel and shoulder devil.
    This shows that, having only my name and age, I have been classified as , ‘heterosexual,’ ‘Muslim’ or ‘A supporter of the regime in Turkey’, since ‘DITIB’ is a controversial institution in Europe, who is doing the Turkish state propaganda under the name of Islam.
    As Crawford discusses, the ads are also a indicator of how stereotypical prejudices can be reproduced through classification.

  6. In modern MENA countries, racial or national segregation is unlikely to pose significant problems for classification systems. The influence of ideological nationalism is waning in many regions due to rapid globalization and the homogenization of human culture. However, a pressing concern is the potential use of facial recognition technology by government officials to suppress opposition. This could hinder efforts to challenge authoritarian regimes and potentially extend their tenure in power. Despite this, the effectiveness of these technologies in combating opposition forces remains uncertain.

  7. KİM: A Controversial AI Tool in Turkey

    A recent video showcasing a facial recognition app named “KİM” (meaning “who” in Turkish) used by Turkish Interior Minister Soylu has ignited controversy. KİM promises identification in just seconds, raising concerns about privacy and ethics.

    Proponents claim KİM’s potential for national security, aiding investigations and locating missing persons. However, critics highlight the violation of data protection laws. Processing biometric data like facial scans requires consent or legal authorization, which KİM seems to lack.

    There are many ethical concerns around the app. Facial recognition can be biased, leading to misidentification, particularly for people of color. The potential for mass surveillance or oppressing the opposition is worrying. Imagine peaceful protestors easily identified and targeted.

    Legally, KİM’s use by the Interior Ministry is questionable. Such applications might fall under intelligence agencies, not the ministry. The lack of transparency surrounding KİM’s development and operation further intensifies concerns about accountability.

    The KİM controversy underscores the need for open discussions about AI-powered identification. A strong legal framework ensuring data privacy and clear guidelines for facial recognition use is essential. Independent oversight and accountability mechanisms are crucial to prevent misuse.

    KİM serves as a reminder of the double-edged sword that is AI. While it offers potential benefits, ethical considerations and legal frameworks must be addressed to safeguard privacy and prevent misuse. Responsible implementation and open dialogue are key to harnessing AI for good.

    For more information on the KİM controversy, you can refer to this BBC Turkish article: https://www.bbc.com/turkce/articles/cd1rmnrjggyo [1].

  8. The “Classification” chapter of the book argues the methods of classifying people and objects of artificial intelligence systems. Crawford challenges these technologies’ historical background and involvement period on their social and political implications of classification. AI systems divide people into categories through complex algorithms using data. Throughout this process, demographic data such as gender, ethnicity, and age are often used; however, it is a concern that these classification systems often reinforce biases. The author argues that the way classification technologies are used, especially in areas such as facial recognition can create surveillance societies. It also draws attention to the fact that these systems can lead to discrimination.

    The development of AI brings various opportunities and risks. The first opportunity of AI is the facial recognition technologies. This can help identify and catch criminals quickly with security cameras and this has the potential to improve public safety. On the other hand, the use of a can create some risks. Facial recognition technologies can lead to individuals being kept under constant surveillance without their consent. This violates the right to privacy of individuals and creates a feeling of living under constant supervision. As a result, the use of AI in facial recognition technologies in the region contains both great opportunities and serious risks. It is of great importance that these technologies are used in an ethical, fair, and transparent manner. To solve the potential problems, these systems must be strongly regulated to protect the privacy rights of individuals, prevent discrimination, and prevent the misuse of technology.

Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert

This site uses Akismet to reduce spam. Learn how your comment data is processed.