05. September 2023 Fabian Lütz
Algorithms increasingly make decisions for humans or support human decision-making (Sunstein 2023). To address cases of discrimination regulatory efforts for Artificial Intelligence (AI) have recently flourished. On 14 July 2023, the UN Human Rights Council adopted Resolution A/HRC/RES/53/29 on New and Emerging Technologies at the 53rd session calling for equal rights offline and online, also regarding gender equality and non-discrimination. It builds on a wide range of texts including the Guiding Principles on Business and Human Rights (UNGPs), the UNESCO Recommendation on the Ethics of AI, which includes a dedicated gender chapter, the Call to Action for Human Rights, the Roadmap for Digital Cooperation, the creation of the Office of the Secretary-General’s Envoy on Technology and the ongoing process of the UN Global Digital Compact.
The present contribution assesses the Resolution through the lens of gender equality, focusing on its added value regarding regulatory efforts for AI by sketching out elements where AI poses opportunities but also risks for gender equality.
Recognising AI’s potential to achieve gender equality …
The Resolution highlights the potential of AI: (i) to facilitate progress on human rights, (ii) to address the digital (gender) divide and to advance gender equality, and (iii) the empowerment of women and girls in line with SDG 5. This positive narrative rebalances the dominant discourse on regulating AI. Although regulation is key to control and mitigate negative impacts on gender equality, the approach should be holistic and consider both positive and negative effects of AI. By referring to the digital (gender) divide the Resolution draws attention to one of the root causes of gender inequalities (Lütz 2023a), biases and algorithmic discrimination. Women are underrepresented in AI professions and contribute less to models and algorithms that shape current and future decision-making. Equally, the digital gender divide hinders access to data (creation) thus limiting women’s role in design and development of AI. The role of AI as “critical enablers of development” (A/HRC/RES/53/29, p. 2) making such technologies available to all is highlighted. These digital divides “may reflect and amplify existing social, cultural and economic inequalities” (A/HRC/RES/53/29, p. 3). Mentioning SDG 5 and gender equality, the Resolution underscores that the achievement of SDG 5 requires a gender lens (Lütz 2023b).
… while raising awareness of its harmful potential
The resolution warns of the risk that AI can affect women disproportionately by perpetuating existing patterns of gender inequality and discrimination which is further exacerbated by the lack of women in AI. It underscores the role of biases that facilitate the perpetuation of existing stereotypes and pave the way for algorithmic discrimination. Such biases need to be addressed in all legal AI frameworks and at technical level (Schwartz et al. 2022). Reference is made to the importance of the equal representation of women and men at the design and coding table for algorithms. The mindset of AI developers shapes and influences the design and functioning of algorithms. Without appropriate guiding principles, it “can entail serious risks to the protection, promotion and enjoyment of human rights, […] in particular by embedding and exacerbating bias which potentially result in discrimination and inequality” (A/HRC/RES/53/29, p. 3).
The HRC’s call for AI regulation that incorporates gender equality and non-discrimination
While acknowledging the difficulties to keep up with the speed of AI development, the Resolution emphasizes that AI systems impacting human rights may require “effective measures to prevent, mitigate and remedy adverse human rights impacts of such technologies” (A/HRC/RES/53/29, p. 3). Human rights law also requires addressing issues of negative stereotyping, which are laid down in Article 5 of CEDAW. Regarding AI regulation, it highlights the “importance of a human rights-based approach […], a holistic understanding of technology and holistic governance and regulatory efforts” (A/HRC/RES/53/29, p. 4). The Resolution also emphasises the responsibility of business to ensure the respect of human rights based on the UNGPs, referring to the potential of human rights due diligence which could contribute to address the risks associated with biases and algorithmic discrimination. The Resolution underscores
“the importance of ensuring appropriate safeguards and human oversight […] and of respecting and promoting human rights in national, regional and international regulatory frameworks and legislation, and on the conception, design, use, development, further deployment and impact assessments and technical standard-setting of new and emerging digital technologies while ensuring the meaningful participation of all relevant stakeholders, including the private sector, academia, the media and civil society” (A/HRC/RES/53/29, p. 4).
Food for thought for AI regulation
The Resolution invites the international community to consider the protection of individuals against harm caused by AI by introducing frameworks for impact assessments, due diligence, effective remedies, human oversight, accountability and legal responsibility. Protecting individuals from gender-based discrimination and promoting transparency of AI are mentioned alongside the strengthening of oversight and enforcement capacity. First, it is sketched out how individuals could be protected from harm caused by AI, notably ensuring AI safety and by “introducing frameworks for impact assessments related to human rights, exercising due diligence to assess, prevent and mitigate adverse human rights impacts, and ensuring effective remedies and human oversight, accountability and legal responsibility” (A/HRC/RES/53/29, p. 4). Various tools have been already proposed by the OECD, UNESCO or CoE in AI soft law frameworks (Lütz 2023a).
If these elements are included in future international or regional legislative frameworks, this would be good news for gender equality. In addition, concrete deliverables are requested by the Resolution to UN entities, namely a report by the Office of the High Commissioner (OHCHR), an expansion of the capacities of OHCHR to advance human rights in the context of AI and a continuation of the work on the practical application of the UNGPS regarding activities of technology companies by OHCHR. These measures would enable OHCHR to continue its important work on AI and human rights and pave the way to ensure an approach that incorporates gender equality in all its actions.
Outlook for global AI regulation
Resolution A/HRC/RES/53/29 needs to be seen in the context of calls for global AI regulation and the creation of AI agency to ensure the safety and the respect of human rights (Guterrez 2023). Academics and the UN High Commissioner for Human Rights have called for AI regulation, preferably at international level and establishing a dedicated AI regulatory agency. The UN Secretary General equally calls for AI regulation and will convene a multistakeholder High-Level Advisory Board for AI (Guterrez 2023).
At regional level, proposals such as the EU Artificial Intelligence Act (EU 2021) and the Council of Europe Draft Framework Convention on Artificial Intelligence look promising and can be expected to be adopted in 2023 (Lütz 2022). The Council of Europe also works on specific rules on AI, gender equality and non-discrimination.
At national level, China and Brazil proposed laws containing specific provisions on gender equality and algorithmic gender-based discrimination. To ensure the protection against discrimination in the offline and online world, international organizations (Council of Europe, OECD, UNESCO) and individual countries (e.g. Canada, USA, Switzerland) have already enacted recommendations on human rights and AI which address some risks of bias and discrimination. However, rather than non-binding rules, it can be hoped that the HRC 53 AI Resolution and the UN’s work on AI will give new impetus to global and regional AI regulation to address both risks and opportunities for GE and non-discrimination. The UN Global Digital Compact might be one avenue to pursue, even though any outcome will be non-binding norms but only policy guidance (Lütz 2023d). However, there is a need for legislative frameworks that tackle biases and gender-based algorithmic discrimination (Lütz 2023c). In this regard, EU efforts or Brazil’s new AI law that specifically addresses the issue of gender-based algorithmic discrimination are a step in the right direction. As AI’s impacts are shaped and felt globally, there is an explicit need for international coordination and regulation. The Resolution can accelerate the inclusion of gender equality and non-discrimination on the regulatory AI agenda.
European Union (2021). Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL LAYING DOWN HARMONISED RULES ON ARTIFICIAL INTELLIGENCE (ARTIFICIAL INTELLIGENCE ACT) AND AMENDING CERTAIN UNION LEGISLATIVE ACTS, COM/2021/206 final, available at: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021PC0206 (date of access: 24 August 2023).
Guterrez, António (2023). UN Secretary-General's remarks to the Security Council on Artificial Intelligence, 18 July 2023, available at: https://www.un.org/sg/en/content/sg/speeches/2023-07-18/secretary-generals-remarks-the-security-council-artificial-intelligence (date of access: 24 August 2023).
Human Rights Council Resolution A/HRC/RES/53/29, New and emerging digital technologies and human rights, 14 July 2023, available at: https://documents-dds-ny.un.org/doc/UNDOC/LTD/G23/140/62/PDF/G2314062.pdf?OpenElement (date of access: 24 August 2023).
Lütz, Fabian (2022). Gender equality and artificial intelligence in Europe. Addressing direct and indirect impacts of algorithms on gender-based discrimination. ERA Forum 23, 33–52. https://doi.org/10.1007/s12027-022-00709-6
Lütz, Fabian (2023a). Algorithmische Entscheidungsfindung aus der Gleichstellungsperspektive – ein Balanceakt zwischen Gender Data Gap, Gender Bias, Machine Bias und Regulierung, GENDER – Zeitschrift für Geschlecht, Kultur und Gesellschaft, 15(1), 26–41. https://doi.org/10.3224/gender.v15i1.03
Lütz, Fabian (2023b). Gender Equality and Artificial Intelligence: SDG 5 and the Role of the UN in Fighting Stereotypes, Biases, and Gender Discrimination. In Elisa Fornalé & Federica Cristani (eds), Women’s Empowerment and Its Limits. Interdisciplinary and Transnational Perspectives Toward Sustainable Progress, pp. 153–180. Cham: Palgrave Macmillan. https://doi.org/10.1007/978-3-031-29332-0_9
Lütz, Fabian (2023c). Le rôle du droit pour contrer la discrimination algorithmique dans le recrutement automatisé. In Florence Guillaume (ed.), La technologie, l‘humain et le droit (Conférences CUSO, pp. 235–258), Bern: Stämpfli.
Lütz, Fabian (2023d). Human Rights, Discrimination, and the Regulation of AI with a special focus on Gender Equality (Contribution for the United Nations Global Digital Compact, Office of the Secretary-General's Envoy on Technology).
Schwartz, Reva; Vassilev, Apostol; Greene, Kristen; Perine, Lori; Burt, Andrew & Hall, Patrick (2022). NIST Special Publication 1270, Towards a Standard for Identifying and Managing Bias in Artificial Intelligence. https://doi.org/10.6028/NIST.SP.1270
Sunstein, Cass (2023). Decisions about Decisions: Practical Reason in Ordinary Life, pp. 159–180, Cambridge: Cambridge University Press. https://doi.org/10.1017/9781009400480
UNESCO Recommendation on the Ethics of Artificial Intelligence, 23 November 2021, available at: https://unesdoc.unesco.org/ark:/48223/pf0000381137 (date of access: 24 August 2023).
United Nations, Secretary General (2020). The Highest Aspiration: A Call to Action for Human Rights. Available at: https://www.un.org/sg/sites/www.un.org.sg/files/atoms/files/The_Highest_Asperation_A_Call_To_Action_For_Human_Right_English.pdf (date of access: 24 August 2023).
United Nations, Report of the Secretary-General, Roadmap for Digital Cooperation, June 2020, available at: https://www.un.org/en/content/digital-cooperation-roadmap/assets/pdf/Roadmap_for_Digital_Cooperation_EN.pdf (date of access: 24 August 2023).
Zitation: Fabian Lütz: Gender Equality and Algorithms: HRC Resolution on New and Emerging Technologies, in: blog interdisziplinäre geschlechterforschung, 05.09.2023, www.gender-blog.de/beitrag/gender-equality-algorithms-hrc-resolution/, DOI: https://doi.org/10.17185/gender/20230905
Beitrag (ohne Headergrafik) lizensiert unter einer Creative Commons Namensnennung 4.0 International Lizenz