Gendered Borders: Unveiling the Impact of EU’s Technological Fortress on Women and Queer Refugees

Temps de lecture : 14 minutes

Rosário Frada

14/11/2023

Emerging Technology at EU Borders

On a global scale, women and queer people are forced to flee their home due to gender-specific persecution. Despite supposed to provide protection, these patterns of discrimination permeate the EU’s asylum system in several forms, meaning refugees fleeing gender-specific violence are frequently forced into a system that is ill-suited to protect them from harm. Indeed, these dangers are exacerbated by the EU’s efforts to automate border control for security purposes. From military-grade drones to experimental technology and Artificial Intelligence (AI) lie-detection software, the EU and its members have invested hundreds of millions of euros on technologies aimed at tracking down and deterring refugees on its borders[1]Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, … Continue reading, explaining labels such as “technological fortress[2]Marin Luisa, ‘Is Europe Turning into a “Technological Fortress”? Innovation and Technology for the Management of EU’s External Borders: Reflections on FRONTEX and EUROSUR’, in Regulating … Continue reading”, and “cyber-fortress[3]Guild Elspeth, Sergio Carrera, and Florian Geyer, ‘The Commission’s New Border Package: Does It Take Us One Step Closer to a “Cyber-Fortress Europe”?’ (Centre for European Policy Studies, … Continue reading”.

Extensive research exists regarding biometric technology for migration management. However, conversations on the potential implications of border technology on gender equality and the enjoyment of other human rights is still lacking. As part of the EU’s intentions to deter irregular migration, in 2019 Hungary, Latvia and Greece launched the EU-funded project iBorderCtrl: an AI-lie detector that, with the help of facial recognition technology, scans migrants’ facial expressions and biomarkers of deceit to identify bona fide[4]Bona fide, or good faith, is one of the fundamental legal principles of Public International Law. It imposes a moral behavioural standard of honesty and loyalty, without fraud or deceit. A bona fide … Continue reading travellers and exclude unwanted migrants and that subsequently transfers that information to border officials[5]Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, … Continue reading. Importantly, note that iBorderCtrl’s functions are grounded on the myopic assumption that “universal ways of expressing deception through non-verbal expressions[6]Javier Sánchez-Monedero and Lina Dencik, ‘The Politics of Deceptive Borders: “Biomarkers of Deceit” and the Case of iBorderCtrl’ 25, no. 3 (2022): 421, … Continue reading” exist, dismissing the complexity behind human behaviour. Whilst ending after facing a lawsuit by MEP Patrick Breyer for violating fundamental human rights[7]Patrick Breyer, ‘EU-Funded Technology Violates Fundamental Rights’, European Voices on Surveillance (blog), 2021,  … Continue reading, the EU continues to advance iBorderCtrl’s technology as an “innovative border management solution[8]Horizon2020, ‘robusT Risk basEd Screening and Alert System for PASSengers and Luggage’, European Commission (blog), n.d., https://cordis.europa.eu/project/id/787120.”.

With this in mind, how does the deployment of digital technologies at EU borders increase gender discrimination and the exclusion of women and queer migrants and refugees? 

This article investigates the ways in which the EU’s turn to technological border control has exacerbated the gender gap in refugee protection regime, subjecting women and queer individuals to further bias and revealing the need for additional protection. It sheds light on the various forms of discrimination they face on the grounds of their gender and proposes ways to engage in conversations regarding EU border technology and gender equality.

An Ill-Suited System For Gender-Specific Persecution: A Feminist Critique of Refugee Law

The 1951 Convention Relating to the Status of Refugees (Refugee Convention) is a United Nations multilateral treaty that defines the term refugee and establishes the rights of those entitled to protection and the duties assigned to the states granting asylum. Signed by every member state of the EU, it requires asylum claimants to demonstrate they meet the refugee definition as per Article 1A(2) as soon as they cross international borders, namely showing they have a “well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion[9]United Nations, 1951 Convention Relating to the Status of Refugees, art. 1A(2).”.

Although Article 1A(2) recognizes multiple forms of persecution and requires a deep investigation into specific features and circumstances of the individual claimant, many criticise the convention for its gender blindness[10]Georgina Firth and Barbara Mauthe, ‘Refugee Law, Gender and the Concept of Personhood’ 25, no. 3 (2013): 470–501, https://doi.org/10.1093/ijrl/eet034., underpinned by the conventional view of male dissidents, imprisoned or tortured by the state due to political activities[11]Nahla Valji, Lee Anne de la Hunt, and Helen Moffett, ‘Where Are the Women? Gender Discrimination in Refugee Policies and Practices’ 55, no. 1 (2003): 61–72.. As claimed by Professor Efrat Arbel, Professor Catherine Dauvergne and Professor Jenni Millbank, refugee jurisprudence is a better fit for men than women since it is inclined towards the experience of public actors, with the classic oppressor being the state[12]Catherine Dauvergne, Efrat Arbel and Jenni Millbank, ‘Gender in Refugee Law: From the Margins to the Centre’ (Routledge, 2016).; the types of violations considered pertain to those “that the men who first articulated the concept most feared[13]Georgina Firth and Barbara Mauthe, ‘Refugee Law, Gender and the Concept of Personhood’ 25, no. 3 (2013): 474, https://doi.org/10.1093/ijrl/eet034.”.

This traditional view of the asylum-seeker may be said to reflect a male paradigm that privileges public activities and failing to recognize that most women activities occur in the private sphere, excluding the recognition of abuse deriving from patriarchal structures and consequently denying women international protection. This is reinforced by the fact that the Refugee Convention’s mindset treats women’s political involvement as personal conduct[14]Valji, de la Hunt, and Moffett, ‘Where Are the Women? Gender Discrimination in Refugee Policies and Practices’.. Subsequently, women are “less likely than men to be seen as subjects of persecution[15]Amy Shuman and Carol Bohmer, ‘Gender and Cultural Silences in the Political Asylum Process’ 17, no. 8 (2014): 939–57, https://doi.org/10.1177/1363460714552262.” by immigration officials, facing more barriers throughout the asylum system which fails to provide a gender-neutral space for more complex, inclusive dialogue between the asylum-seeker, her representative and the one making decisions. Dismissing the uniquely female experience weakens the de facto enjoyment of the rights owed to refugee women; however, note that a gender-sensitive interpretation of the law should not replicate binary structures nor essentialize female refugees.

Furthermore, women applicants are commonly deemed “less credible[16]Merryn McKinnon and Christine O’Connell, ‘Perceptions of Stereotypes Applied to Women Who Publicly Communicate Their STEM Work’ 7, no. 1 (2020): 6.” than their male counterparts, generating a systematic disbelief that has been termed by scholars as the “gender credibility gap[17]Judi Brownell, ‘Communicating with Credibility: The Gender Gap’ 34, no. 2 (1993): 52–61, https://doi.org/10.1016/0010-8804(93)90107-T.”. In fact, the law does not include enough safeguards to ensure the effective exercise of procedural rights, with the adjudication of both the subjective and objective aspects of the application for international protection substantially obstructing women’s right to asylum. For instance, it is extremely hard to corroborate sexual or domestic violence with documentary evidence, creating a form of double violence or double discrimination. Although this is one of the greatest reasons forcing women to flee their countries, caseworkers are provided with wide margins of discretion to assess asylum stories[18]Catherine Briddick, ‘Rethinking Refuge From Gender-Based Violence: Persecution for Which Convention Reason?’, Rethinking Refuge (blog), n.d., … Continue reading.

When initiating their asylum procedure, dark-skinned women immediately face systemic violence and further oppression. Being perceived as “less reliable[19]Amanda Carlin, ‘The Courtroom as White Space: Racial Performance as Noncredibility’, UCLA Law Review (blog), 2009, … Continue reading” implies that at the intersection of racism, sexism and xenophobia, refugee women face a disproportionately high burden of proof. Subsequently, accessing material justice becomes harder and the protection gap widens.

Queer Asylum-Seekers: Navigating a Hetero-Normative Legal Framework

It too seems that queer asylum-seekers are forced to navigate a system that fails to recognize their specific experiences. Across Europe, people seeking asylum on the grounds of gender identity and sexual orientation are routinely having their claims rejected due to a widespread “culture of disbelief[20]Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, … Continue reading” that places the burden of proof solely on the asylum applicant despite international refugee law establishing this is to be shared equally with decision makers[21]United Nations High Commissioner for Refugees, ‘Handbook on Procedures and Criteria for Determining Refugee Status and Guidelines on International Protection’ (United Nations High Commissioner … Continue reading. Sitting within a hostile environment to immigration, one in every queer asylum claim is refused because officials do not believe exiles’ sexual orientation or gender identity[22]Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, … Continue reading. As they are less likely than other asylum-seekers to carry evidence proving their claim, it becomes easier for officials to overturn their applications.

Rather, studies demonstrate Queer claimants must reveal their sexual orientation or gender identity if they are to present themselves as a minority, but they must do so in a way that is credible and legible to authorities[23]Kahn Sarilee and Edward Alessi, ‘Coming Out Under the Gun: Exploring the Psychological Dimensions of Seeking Refugee Status for LGBT Claimants in Canada’ 31, no. 1 (2017): 22–41, … Continue reading. For example, an advocate for lesbian asylum-seekers notes she coaches her clients to act stereotypically masculine due to the assumption that feminine-presenting lesbians are inauthentic, embedding performativity within the asylum system if protection is to be negotiated and attained[24]Elif Sari, ‘Lesbian Refugees in Transit: The Making of Authenticity and Legitimacy in Turkey’ 24, no. 2 (n.d.): 140–58, https://doi.org/10.1080/10894160.2019.1622933.. Paradoxically, however, several asylum-seekers have also been denied protection due to their inability to behave discreetly regarding their non-conforming identity in attempts to avoid persecution[25]Johannes Gartner, ‘(In)Credibly Queer: Sexuality-Based Asylum in the European Union’, Humanity in Action (blog), n.d., … Continue reading. Thus, claimants found to be persecuted frequently arrive at equally persecutory environments where they are forced to adopt a “submissive and compliant behaviour[26]Johannes Gartner, ‘(In)Credibly Queer: Sexuality-Based Asylum in the European Union’, Humanity in Action (blog), n.d., … Continue reading”.

Equally as worrying is the fact that throughout EU countries including Ireland, Austria, Netherlands and Belgium, caseworkers have asked queer applicants for explicit information and evidence of their sexual life during asylum interviews, reducing queer identities to sexual conduct[27]Sabine Jensen and Thomas Spijkerboer, ‘Fleeing Homophobia, Asylum Claims Related to Sexual Orientation and Gender Identity in Europe’ (Vrije Universiteit Amsterdam (VU University Amsterdam), … Continue reading. Essentializing a queer identity to something exclusively behavioural or sexual as opposed to a highly nuanced matter that is central to one’s identity may oblige queer asylum-seekers to perform their identities, embodying expectations of the host State regarding queer communities.

The Gendered and Racialized Premises of Border Technology

The very foundational question regarding gender discrimination in technology and its effects on refugee women is grounded in the contemporary state of research regarding algorithmic bias and the bias of coders themselves. Women are often poorly represented in datasets and thus border algorithms have little possibility to train the matching of female faces to ensure fair representation. In fact, research suggests three out of five immigration officers make biased decisions when possessing previous knowledge regarding the asylum case they have been appointed to[28]Gavett Gretchen, ‘Can Unconscious Bias Undermine Fingerprint Analysis?’, Frontline (blog), 2012, https://www.pbs.org/wgbh/frontline/article/can-unconscious-bias-undermine-fingerprint-analysis/.. Significantly, note that before the development of virtual borders, the only personal data stored in the Eurodac database has been applicants’ fingerprints and gender, making evident that gender considerations might predetermine asylum decisions and drive gendered discrimination[29]‘Regulation (EU) No 603/2013’, § Art 14 (2013).. The fact that human bias is being transformed via data only serves to ascertain that biometric being used to automate asylum may be one of the most devastating sources of discrimination for women in migration.

Despite evidence that stereotypes and racist tropes are embedded in facial recognition technology, the EU-funded lie-detection software has been predominantly tested in white European men[30]Adil Habib, ‘The Ongoing Digitisation of Europe’s Borders’, Digital Freedom Fund (blog), 2021, https://digitalfreedomfund.org/the-ongoing-digitisation-of-europes-borders/., making clear that racial premises too are rooted in AI systems[31]Stephen Buranyi, ‘Rise of the Racist Robots – How AI Is Learning All Our Worst Impulses’, The Guardian (blog), 2017, … Continue reading.This raises additional important questions, including whether border technology is capable of considering cultural and ethnic divergences in body language, important given the diversity of the refugee population arriving at EU borders. Taken together, this would mean that black women seeking asylum in the EU would be considered a suspect by the State long before they were to arrive at any physical checkpoint, rendering their safety to be controlled entirely by technology.

Professor Joseph Pugliese states facial recognition technology employed in border control systems are “infrastructurally calibrated to whiteness[32]Joseph Pugliese, ‘Biometrics, Infrastructural Whiteness and the Zero Degree of Non-Representation’ 34, no. 2 (2007): 107.”; the failure to achieve race neutrality becomes the rhetoric that configures AI. Studies prove the existence of gender and racial bias, with facial recognition software demonstrating an error rate of 34.7% for dark-skinned women in comparison with 0.8% for light-skinned males[33]Larry Hardesty, ‘Study Finds Gender and Skin-Type Bias in Commercial Artificial-Intelligence Systems’, Massachusetts Institute of Technology (blog), 2018, … Continue reading. In fact, AI software developed by leading companies in AI research has generated serious errors, including a Google image recognition program that labelled the faces of several black people as gorillas[34]Stephen Buranyi, ‘Rise of the Racist Robots – How AI Is Learning All Our Worst Impulses’, The Guardian (blog), 2017, … Continue reading. Additionally, a New Zealand machine that deployed the same facial recognition technology as iBorderCtrl rejected pictures of an Asian descent man because the machine failed to detect an open set of eyes[35]James Griffiths, ‘New Zealand Passport Robot Thinks This Asian Man’s Eyes Are Closed’, CNN (blog), 2016, https://edition.cnn.com/2016/12/07/asia/new-zealand-passport-robot-asian-trnd/index.html.. While technological solutions can be drawn, these examples seek to hint at the larger problem: automated border control could drive the replication of large-scale systemic biases that people have spent decades campaigning to combat.

As border control AI will increasingly play a critical role in credibility evaluations, it becomes important to consider the intersection of gender with race and other identity vectors, especially as risk is excessively coupled with non-white, male, heterosexual and cis-gendered persons[36]Stephanie Silverman and Esra Kaytaz, ‘Examining the “National Risk Assessment for Detention” Process: An Intersectional Analysis of Detaining “Dangerousness” in Canada’, 2020, 693–709, … Continue reading. Therefore, will non-white individuals see their asylum applications considered fairly? Note that many asylum-seeking women in Europe are dark-skinned; if 34.7% of them are incorrectly identified, multiple fundamental rights are violated and security issues may arise[37]Larry Hardesty, ‘Study Finds Gender and Skin-Type Bias in Commercial Artificial-Intelligence Systems’, Massachusetts Institute of Technology (blog), 2018, … Continue reading.

Having a non-white-male identity becomes a barrier of entry, enacting dynamics of expulsion and exclusion that oppress women seeking protection. Through lie-detection software, borders become a site where non-white female bodies are barred out, arranging people into racialized and gendered risk categories, reconstructing and objectifying women of colour, hindering access to asylum, and ultimately giving in to the logic of a superior white EU. Whilst AI and authorities are subject to equal decision-making issues, forced migration may not be reduced to an algorithm disregarding human rights. Otherwise, border technology may encourage asylum-seekers to travel through more irregular means, which not only places them in further danger but is then instrumentalized by authorities to define asylum claims as “abusive[38]Violeta Moreno-Lax, Accessing Asylum in Europe. (Oxford: Oxford University Press, 2017), 5.”.

Automating Discrimination Against Queer People

Facial recognition technologies employed by iBorderCtrl may automate Queer oppression, reducing one’s ability to define one’s gender identity. Research shows AI is frequently trained to believe gender is “binary, immutable[39]OS Keyes, ‘The Misgendering Machines: Trans/HCI Implications of Automatic Gender Recognition’, 2018, 88, https://doi.org/10.1145/3274357.”, excluding gender non-conforming and trans individuals from its architecture and identifying an intellectual failure in border control practices that must be addressed. Misgendering AI technologies not only make one feel unseen, deteriorating trans peoples’ physical and mental health and increasing the likelihood that algorithmic binary logics will harm gender non-conforming migrants, but fail to capture objective data[40]Sabra Katz-Wise, ‘Misgendering: What It Is and Why It Matters’, Harvard Health Publishing, 2021, https://www.health.harvard.edu/blog/misgendering-what-it-is-and-why-it-matters-202107232553.. Professor Eric D. Olson and Professor Kelly Reddy-Best show body-scanning systems miscategorise Queer individuals[41]Eric Olson and Kelly Reddy-Best, ‘“Pre-Topsurgery, the Body Scanning Machine Would Most Likely Error:” Transgender and Gender Nonconforming Travel and Tourism Experiences’ 70, no. 1 (2019): … Continue reading. If lie-detection software believes they are falsifying themselves, they could be considered terrorists and face criminal law.

Subsequently, the notion of the state/private agency as a protector and the public as protected becomes rooted in gender when facial recognition technologies are employed by the EU’s border control practices[42]Laura Shepherd and Laura Sjoberg, ‘Trans- Bodies in/of War(s): Cisprivilege and Contemporary Security Strategy’ 101, no. 1 (2012): 5–23.. In the conventional understanding of the state’s role, there is an implicit assumption of impartiality and commitment to ensuring the safety and well-being of all. However, the incorporation of AI systems trained on binary and immutable notions of gender in border control disrupts this equilibrium, instead reinforcing and perpetuating binary gender norms and creating a system that inadvertently protects some individuals at the expense of others. Ultimately, equating transgenders to fraudsters propagates cis-privilege and cissexist power structures, making Queer asylum-seeker identities invisible in politics and aggravating their vulnerability.

This generates two options that deny freedom of expression and personhood development: revealing “who one is” and being labelled a deceiver who instrumentalizes gender and sexuality to obtain protection, or declining to do so but being exposed as a security threat by border technology[43]Talia Bettcher, ‘Evil Deceivers and Make-Believers: On Transphobic Violence and the Politics of Illusion.’ 22, no. 3 (2007): 43–65, https://doi.org/10.1111/j.1527-2001.2007.tb01090.x.. Nevertheless, in Appellants S395/2002 and S396/2002 v. Minister for Immigration and Multicultural Affairs, the Australian High Court rejected the notion that asylum-seekers could conceal their sexuality for their own protection[44]Jenni Millbank, ‘From Discretion to Disbelief: Recent Trends in Refugee Determinations on the Basis of Sexual Orientation in Australia and the United Kingdom’ 13, no. 2–3 (2010): 391–414, … Continue reading. Similarly in Austria, an Afghan teenager claimed he did not immediately disclose his sexual orientation because he felt intimidated by his translator but this failed to convince asylum officers[45]​​Rick Noack, ‘Afghan Teenager Loses Austrian Asylum Case for Not Acting Gay Enough’, Washington Post (blog), 2018, … Continue reading. Evidently, if this pattern continues and information is used by asylum authorities, oppressive structures may force asylum-seekers to disclose intimate information. Hence, digital borders deconstruct the inalienability, indivisibility and interdependence of human rights, making the right to seek asylum and the right to freedom of expression mutually exclusive. Their security rationale becomes a form of discursive violence, denying the normalisation and emancipation of gender non-conforming and trans people and failing to account for (trans)corporeality.

Through border technology, the body personifies discriminatory border regimes. Asylum claims based on social groups like Queer are usually discredited[46]Jenni Millbank and Laurie Berg, ‘Constructing the Personal Narratives of Lesbian. Gay and Bisexual Asylum Claimants’ 22, no. 2 (2009): 195–223, https://ssrn.com/abstract=1339581.. Multi-layered barriers to asylum arise; Queer asylum-seekers not only flee persecution and face being discredited but now must demonstrate credibility by disclosing personal information to an impersonal border control system despite potentially being used to hide their identities in their countries of origin/residency to safeguard protection. Oppressing sexual minorities to impose heteronormativity is persuaded by various governments that have national laws criminalising same-sex relations including Afghanistan, Kuwait and Sierra Leone; asylum-seekers are likely to find iBorderCtrl and border officers equally threatening[47]#Outlawed: “The Love That Dare Not Speak Its Name”’, Human Rights Watch (blog), n.d., https://features.hrw.org/features/features/lgbt_laws/. This raises the questions: does iBorderCtrl consider lying to protect? Can lying be legitimate if it guarantees one’s safety? Paradoxically, lies that have safeguarded one’s life can place one back in danger. Being believed plausible depends on sexualised and gendered suppositions, failing to uphold the right to non-discrimination and the recognition of one’s equal dignity. Sexual and gender orientation politics’ expansion becomes crucial if human rights are to drive social and institutional change in migration governance, redefining relations between traditional and contemporary challenges.

Lastly, studies show that women and the queer community are more likely to become victims of violence emerging from patriarchal structures, culture and socio-economic status before, during and after their displacement journey, being subjected to traumatic experiences that cause several mental health challenges[48]Konstantina Davaki, ‘The Traumas Endured by Refugee Women and Their Consequences for Integration and Participation in the EU Host Country’, 2021, … Continue reading. Nevertheless, Professor Petra Molnar demonstrates concerns about AI failing to recognise trauma and its impact on memory; trauma survivors’ micro-facial expressions/gestures may be wrongly perceived as deceit by lie-detection software such as iBorderCtrl[49]Petra Molnar, ‘Emerging Voices: Immigration, Iris-Scanning and iBorderCTRL–The Human Rights Impacts of Technological Experiments in Migration’, Opinio Juris (blog), n.d., … Continue reading. Asylum-seekers are unlikely to possess torture or detention evidence; decision-making becomes reliant on a credibility evaluation of trauma history, for which iBorderCtrl is not prepared[50]Jane Herlihy and Stuart Turner, ‘Untested Assumptions: Psychological Research and Credibility Assessment in Legal Decision-Making’ 6, no. 1 (2015): 27380, https://doi.org/10.3402/ejpt.v6.27380.. Furthermore, memory activation conveys a sense of current threat which may stop a trauma survivor from providing a coherent narrative of events, emphasising the danger of presuming one can talk about traumatic experiences[51]Katy Robjant and Mina Fazel, ‘The Emerging Evidence for Narrative Exposure Therapy: A Review’ 30, no. 8 (2010): 1030–39, https://doi.org/10.1016/j.cpr.2010.07.004.. The fact that trauma is gendered and raced, combined with the fact that migrants’ gender identity is likely to be discriminated against by AI lie-detection, raises immense ethical concerns and forces refugees to confront violent obstacles if they are to come into contact with lie-detection border technology.

Advancing Gender Equality in an Increasingly Digitizing World

The automation of suspicion towards asylum-seekers exemplifies the EU’s race to AI through an industrial expansion aimed at tightening migration governance and capitalising on preventing women and queer individuals from pursuing their human right to seek asylum. This article has proven that iBorderCtrl’s potential lack of neutrality creates invisible borders which replicate existing structural biases and colonial forms of power that subject women and queer people to further violence and forces their bodies into invisibility. This raises serious ethical, legal and political concerns about the gendered, sexualized and raced ways in which lie-detection software defines people worthy of protection, especially considering asylum-seekers’ intricate identities.

Invisible borders must be revised by diversifying the data fed into AI systems and producing assessments that evaluate its impacts. For this, a global cooperation system for mutual human development requires including migration scholars, asylum-seekers and organisations representing them in the dialogue. Simultaneously, border officials must be trained to gain awareness on how sexual orientation and gender identity constitute viable grounds for seeking asylum[52]Vítor Lopes Andrade et al., ‘Queering Asylum in Europe: A Survey Report’ (University of Sussex, 2020), … Continue reading.

Furthermore, existing technology dismisses the unique experiences of women and queer refugees, requiring a gender-sensitive approach if it is to take a human-centric approach to refugees and avoid perpetuating the invisibility of gender-based violence. However, this must be done in a way that avoids replicating binary structures and treating gender as a distinct characteristic from the mainstream refugee categories. This approach to technology and the refugee regime must not essentialize women and queer people but instead create a space for nuanced conversations in which material justice may be equally accessed by all.

One must carefully ponder whether raising ethical implications efficiently combats AI’s tendency to violate human rights. Human rights law is internationally binding, potentially allowing responsibility to be attributed and permitting refugees to access their right to redress. Whilst this framework would not encompass non-signatories, to include a category that would allow for inferences regarding participating agents could potentially embrace state and non-state actors, especially if combined with an adequate accountability framework. If coupled with context-specific oversight systems that acknowledge the displacement experience, the possibility to shape algorithmic accountability and empower asylum-seekers would work towards a human-centred world order.

Despite such recommendations for risk mitigation against gender discrimination, fundamental questions worthy of dedicated investigation remain: are these measures sufficient to overcome gendered, racialized and sexualized logics if not combined with broader cultural and social change? It seems that the refugee regime may only truly become just towards women and the queer community when wider structural reforms that displace discriminatory attitudes take place. Lastly, why is the EU channelling all its efforts to developing AI lie detectors that exclude refugees, when it could be deploying them to rule out gender discrimination at its borders? Only by reflecting upon these matters may the EU guarantee women and queer people are able to pursue a life in safety and with dignity.

To quote this article : Rosário Frada. (2023). Gendered Borders: Unveiling the Impact of EU’s Technological Fortress on Women and Queer Refugees. Gender in Geopolitics Institute. https://igg-geo.org/?p=16383&lang=en 

The statements in this article are the sole responsibility of the author.

References

References
1 Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, https://www.theguardian.com/global-development/2021/dec/06/fortress-europe-the-millions-spent-on-military-grade-tech-to-deter-refugees.
2 Marin Luisa, ‘Is Europe Turning into a “Technological Fortress”? Innovation and Technology for the Management of EU’s External Borders: Reflections on FRONTEX and EUROSUR’, in Regulating Technological Innovation (Palgrave Macmillan, 2011), 131.
3 Guild Elspeth, Sergio Carrera, and Florian Geyer, ‘The Commission’s New Border Package: Does It Take Us One Step Closer to a “Cyber-Fortress Europe”?’ (Centre for European Policy Studies, 2008), https://cdn.ceps.eu/wp-content/uploads/2009/08/1622.pdf.
4 Bona fide, or good faith, is one of the fundamental legal principles of Public International Law. It imposes a moral behavioural standard of honesty and loyalty, without fraud or deceit. A bona fide refugee is thus one who flees life-threatening persecution. However, many criticise this category, claiming it to be a legal and political fiction created to support governments in their attempts to circumvent the duty to protect.
5, 20 Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, https://www.theguardian.com/global-development/2021/dec/06/fortress-europe-the-millions-spent-on-military-grade-tech-to-deter-refugees.
6 Javier Sánchez-Monedero and Lina Dencik, ‘The Politics of Deceptive Borders: “Biomarkers of Deceit” and the Case of iBorderCtrl’ 25, no. 3 (2022): 421, https://doi.org/10.1080/1369118X.2020.1792530.
7 Patrick Breyer, ‘EU-Funded Technology Violates Fundamental Rights’, European Voices on Surveillance (blog), 2021,  https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwjK2dTErLmCAxVAUaQEHZazDTEQFnoECBgQAQ&url=https%3A%2F%2Faboutintel.eu%2Ftransparency-lawsuit-iborderctrl%2F&usg=AOvVaw2AfDIkWI2bTqmaT_O2-cKW&opi=89978449.
8 Horizon2020, ‘robusT Risk basEd Screening and Alert System for PASSengers and Luggage’, European Commission (blog), n.d., https://cordis.europa.eu/project/id/787120.
9 United Nations, 1951 Convention Relating to the Status of Refugees, art. 1A(2).
10 Georgina Firth and Barbara Mauthe, ‘Refugee Law, Gender and the Concept of Personhood’ 25, no. 3 (2013): 470–501, https://doi.org/10.1093/ijrl/eet034.
11 Nahla Valji, Lee Anne de la Hunt, and Helen Moffett, ‘Where Are the Women? Gender Discrimination in Refugee Policies and Practices’ 55, no. 1 (2003): 61–72.
12 Catherine Dauvergne, Efrat Arbel and Jenni Millbank, ‘Gender in Refugee Law: From the Margins to the Centre’ (Routledge, 2016).
13 Georgina Firth and Barbara Mauthe, ‘Refugee Law, Gender and the Concept of Personhood’ 25, no. 3 (2013): 474, https://doi.org/10.1093/ijrl/eet034.
14 Valji, de la Hunt, and Moffett, ‘Where Are the Women? Gender Discrimination in Refugee Policies and Practices’.
15 Amy Shuman and Carol Bohmer, ‘Gender and Cultural Silences in the Political Asylum Process’ 17, no. 8 (2014): 939–57, https://doi.org/10.1177/1363460714552262.
16 Merryn McKinnon and Christine O’Connell, ‘Perceptions of Stereotypes Applied to Women Who Publicly Communicate Their STEM Work’ 7, no. 1 (2020): 6.
17 Judi Brownell, ‘Communicating with Credibility: The Gender Gap’ 34, no. 2 (1993): 52–61, https://doi.org/10.1016/0010-8804(93)90107-T.
18 Catherine Briddick, ‘Rethinking Refuge From Gender-Based Violence: Persecution for Which Convention Reason?’, Rethinking Refuge (blog), n.d., https://www.rethinkingrefuge.org/articles/rethinking-refuge-from-gender-based-violence-persecution-for-which-conventi .
19 Amanda Carlin, ‘The Courtroom as White Space: Racial Performance as Noncredibility’, UCLA Law Review (blog), 2009, https://www.uclalawreview.org/wp-content/uploads/2019/09/Carlin-final-article-no-bleed.pdf .
21 United Nations High Commissioner for Refugees, ‘Handbook on Procedures and Criteria for Determining Refugee Status and Guidelines on International Protection’ (United Nations High Commissioner for Refugees, 2019), https://www.unhcr.org/sites/default/files/legacy-pdf/5ddfcdc47.pdf.
22 Kaamil Ahmed and Lorenzo Tondo, ‘Fortress Europe: The Millions Spent on Military-Grade Tech to Deter Refugees’, The Guardian (blog), 2021, https://www.theguardian.com/global-development/2021/dec/06/fortress-europe-the-millions-spent-on-military-grade-tech-to-deter-refugees.
23 Kahn Sarilee and Edward Alessi, ‘Coming Out Under the Gun: Exploring the Psychological Dimensions of Seeking Refugee Status for LGBT Claimants in Canada’ 31, no. 1 (2017): 22–41, https://doi.org/10.1093/jrs/fex019.
24 Elif Sari, ‘Lesbian Refugees in Transit: The Making of Authenticity and Legitimacy in Turkey’ 24, no. 2 (n.d.): 140–58, https://doi.org/10.1080/10894160.2019.1622933.
25, 26 Johannes Gartner, ‘(In)Credibly Queer: Sexuality-Based Asylum in the European Union’, Humanity in Action (blog), n.d., https://humanityinaction.org/knowledge_detail/incredibly-queer-sexuality-based-asylum-in-the-european-union/.
27 Sabine Jensen and Thomas Spijkerboer, ‘Fleeing Homophobia, Asylum Claims Related to Sexual Orientation and Gender Identity in Europe’ (Vrije Universiteit Amsterdam (VU University Amsterdam), 2011), https://www.refworld.org/docid/4ebba7852.html
28 Gavett Gretchen, ‘Can Unconscious Bias Undermine Fingerprint Analysis?’, Frontline (blog), 2012, https://www.pbs.org/wgbh/frontline/article/can-unconscious-bias-undermine-fingerprint-analysis/.
29 ‘Regulation (EU) No 603/2013’, § Art 14 (2013).
30 Adil Habib, ‘The Ongoing Digitisation of Europe’s Borders’, Digital Freedom Fund (blog), 2021, https://digitalfreedomfund.org/the-ongoing-digitisation-of-europes-borders/.
31, 34 Stephen Buranyi, ‘Rise of the Racist Robots – How AI Is Learning All Our Worst Impulses’, The Guardian (blog), 2017, https://www.theguardian.com/inequality/2017/aug/08/rise-of-the-racist-robots-how-ai-is-learning-all-our-worst-impulses.
32 Joseph Pugliese, ‘Biometrics, Infrastructural Whiteness and the Zero Degree of Non-Representation’ 34, no. 2 (2007): 107.
33, 37 Larry Hardesty, ‘Study Finds Gender and Skin-Type Bias in Commercial Artificial-Intelligence Systems’, Massachusetts Institute of Technology (blog), 2018, https://news.mit.edu/2018/study-finds-gender-skin-type-bias-artificial-intelligence-systems-0212.
35 James Griffiths, ‘New Zealand Passport Robot Thinks This Asian Man’s Eyes Are Closed’, CNN (blog), 2016, https://edition.cnn.com/2016/12/07/asia/new-zealand-passport-robot-asian-trnd/index.html.
36 Stephanie Silverman and Esra Kaytaz, ‘Examining the “National Risk Assessment for Detention” Process: An Intersectional Analysis of Detaining “Dangerousness” in Canada’, 2020, 693–709, https://doi.org/10.1080/1369183X.2020.1841613.
38 Violeta Moreno-Lax, Accessing Asylum in Europe. (Oxford: Oxford University Press, 2017), 5.
39 OS Keyes, ‘The Misgendering Machines: Trans/HCI Implications of Automatic Gender Recognition’, 2018, 88, https://doi.org/10.1145/3274357.
40 Sabra Katz-Wise, ‘Misgendering: What It Is and Why It Matters’, Harvard Health Publishing, 2021, https://www.health.harvard.edu/blog/misgendering-what-it-is-and-why-it-matters-202107232553.
41 Eric Olson and Kelly Reddy-Best, ‘“Pre-Topsurgery, the Body Scanning Machine Would Most Likely Error:” Transgender and Gender Nonconforming Travel and Tourism Experiences’ 70, no. 1 (2019): 250–61, https://doi.org/10.1016/j.tourman.2018.08.024.
42 Laura Shepherd and Laura Sjoberg, ‘Trans- Bodies in/of War(s): Cisprivilege and Contemporary Security Strategy’ 101, no. 1 (2012): 5–23.
43 Talia Bettcher, ‘Evil Deceivers and Make-Believers: On Transphobic Violence and the Politics of Illusion.’ 22, no. 3 (2007): 43–65, https://doi.org/10.1111/j.1527-2001.2007.tb01090.x.
44 Jenni Millbank, ‘From Discretion to Disbelief: Recent Trends in Refugee Determinations on the Basis of Sexual Orientation in Australia and the United Kingdom’ 13, no. 2–3 (2010): 391–414, https://doi.org/10.1080/13642980902758218.
45 ​​Rick Noack, ‘Afghan Teenager Loses Austrian Asylum Case for Not Acting Gay Enough’, Washington Post (blog), 2018, https://www.washingtonpost.com/world/2018/08/16/afghan-teenager-loses-austrian-asylum-case-not-acting-gay-enough/.
46 Jenni Millbank and Laurie Berg, ‘Constructing the Personal Narratives of Lesbian. Gay and Bisexual Asylum Claimants’ 22, no. 2 (2009): 195–223, https://ssrn.com/abstract=1339581.
47 #Outlawed: “The Love That Dare Not Speak Its Name”’, Human Rights Watch (blog), n.d., https://features.hrw.org/features/features/lgbt_laws/
48 Konstantina Davaki, ‘The Traumas Endured by Refugee Women and Their Consequences for Integration and Participation in the EU Host Country’, 2021, https://rm.coe.int/ipol-stu-2021-691875-en-1-/1680a23902.
49 Petra Molnar, ‘Emerging Voices: Immigration, Iris-Scanning and iBorderCTRL–The Human Rights Impacts of Technological Experiments in Migration’, Opinio Juris (blog), n.d., http://opiniojuris.org/2019/08/19/emerging-voices-immigration-iris-scanning-and-iborderctrl-the-human-rights-impacts-of-technological-experiments-in-migration/.
50 Jane Herlihy and Stuart Turner, ‘Untested Assumptions: Psychological Research and Credibility Assessment in Legal Decision-Making’ 6, no. 1 (2015): 27380, https://doi.org/10.3402/ejpt.v6.27380.
51 Katy Robjant and Mina Fazel, ‘The Emerging Evidence for Narrative Exposure Therapy: A Review’ 30, no. 8 (2010): 1030–39, https://doi.org/10.1016/j.cpr.2010.07.004.
52 Vítor Lopes Andrade et al., ‘Queering Asylum in Europe: A Survey Report’ (University of Sussex, 2020), https://www.sogica.org/wp-content/uploads/2020/07/The-SOGICA-surveys-report_1-July-2020-1.pdf.