Queering the law: How to make AI with a Queer Perspective from the Global Majority

Umut Pajaro Velasquez

Funding Area: Community governance / Gobernanza comunitaria

Problem Statement: The way to understand and define "intelligence" respond to a specific way of knowing. AI reproduces this, and queer and critical theories have shown how it excludes others, including those traditionally considered feminine, and this intelligence has its equivalence to white man's knowledge. AI thus engages in a sociotechnical exclusion or repression of women's knowledge, reifying a gender- and race-based conceptualization of "intelligence". Also, AI may offer an opportunity to change assumptions about male and female epistemology. For example, narratives of "hard" and "soft" intelligence are often classified as male and female, respectively. Adrian Weller points out that this "hard" intelligence, which encompasses logic and rationality, is easier to reproduce in technological form, but "soft" intelligence and creative problem-solving, empathy, negotiation, and persuasion can become privileged as it is difficult to codify. Whether AI is thought to depend on embodying a male epistemology, or whether AI promises to give a feminist epistemology a head start, AI is perpetuating and reinforcing binary gender stereotypes, leaving behind those who are outside the normative and binary concepts of gender. Hence, we propose a queer perspective for AI in order to create an common framework in all AI life-cycle with a manly focus on the Global Majority.

Proposed Activities: In order to get this activity done we need to follow the next steps: 1. Collection of Data: Old ways for new technologies, Bottom – Up, Look, Listen, Write 2. Analysis of Data: Mixture of techniques and tools, Transdisciplinary and multistakeholder approach, 3. Publication of Data: Tell their story, they aren’t just data. 4. Openness: Found the language to unify policy, academia, real life and technical standards with the less “lost in the translation”. Create a report that can invite others to contribute to the research’s next steps. 5. Creation of the Comunity: A repository on GitHub with the description of the project or a webpage that allows me to create a community and also bring new gendered perspectives from the Global Majority and how to include on the law and technical standard with the less biased outcome possible Find new ways to create meaningful participation in all stages of the AI cycle and also in the policymaking process for people outside the gender-binary.

Openness: In the case, I'm describing having a more open and transparent way of getting data is the best way to find that common language that I want for this future framework because I believe that only with collaborative knowledge is the way we are going to find the solutions that humanity is facing right now, especially those related to restrictions of human rights. That’s why after the firsts explorative and descriptive research with different communities I will be sharing though a GitHub page all the steps I follow to get my conclusions so anyone anywhere can replicate them and add some new information to this new develop grid of genders from the Global Majority. Allowing to this new community to think as a next step on how this sistematization can help to get new ways to create meaningful participation in all stages of the AI cycle and also in the policymaking process for people outside the gender-binary. But beyond all of these reasons having an open infrastucture for projects like will have by default a Human-Centered AI because it will guarantee Accessibility and Inclusivity, Transparency and Accountability, Ethics and Fairness, Collaboration and Knowledge Sharing, Community Engagement and Empowerment, and address biases and discrimination.

Challenges: 1. The lack of representation and embedded codification of the different and diverse genders that are part of the Global Majority into the development of new policies and technologies such as AI. 2. My main challenge until now with working with projects that are open by default is to know until which point the result, the recollection of the data, and concede the ability to replicate that project as it was done should be open and shared, I explain myself, the challenge is on which conditions, terms the project is going to be open and how people that are to use, replicate or are part of it going to acknowledge who are the authors. 3. Found as a first requirement a population where gender is not seems as binary, then data collection resources, and contributors to build the first draft of the framework. 4. How to create a community who wishes to undertake an open AI project within a framework of gender–diversity on the global majority where this topic is still controversial.

Neglectedness: Never applied before to finance this project because before of being part of the OLS fellowship I don’t even think about on the possibility on doing so, but now I know so that’s why I’m here doing my first application for funds for a project that I’d been self-funding until now.

Success: This project in order to measure its success will have the following KPIs: 1. The number of genders, expression of genders, representation of genders, gender social roles, cultures, name or concepts propose as gender markers found with help of the gender-diverse communities. 2. Openness of the research on terms of data protection considering it is a vulnerable and persecuted population. 3. The number of new tools or instruments to collect and present results that came from open data 4. Create an open data and theoretical prototype on how to create a common language that allows anybody to conversation and use in a way that can improve AI models from its foundation in order to get a common good of queer people from the Global Majority. 5. Launch the first prototype 6. Create a community where everyone who wishes to undertake an open AI project within a framework of gender–diversity on the global majority

Total Budget: 6477

Affiliations: Independent

LMIE Carveout: The persons in the charge is from LMIEs country, Colombia, but not only for that reason most of the people involved as volunteers or subjects in this research are xpected to be from the Global Majority that means that most of all the people that is going to take part in some stage are from countries that are listed as LMIEs. We just need to remember that the aim of this project is to create commom language with a queer perspective from the Global Majority for social sciences, policy-making and technical community when it comes to all the AI life-cycle.

Team Skills: The person behind this project not only have knowledge on social sciences investigations tools, methods, but also has skills and expertise in technical capabilities related to AI, and also possesses knowledge in the fields of internet governance and digital policy, allowing them to effectively curate and analyze the contributions made through the Atlas. Also they have ability to manage multiple projects simultaneously and coordinate across different time zones and cultures. They have established relationships with key organizations and networks within the Artificial Intelligence community. These connections enable them to tap into existing resources and leverage the strengths of other initiatives working on similar issues. Also, have participated in international meetings and events focused on digital rights, open source, Artificial Intelligence, Internet governance, bringing firsthand understanding of the challenges facing during the AI life-cycle if you belong to a minority such as queer and BIPOC people. Furthermore, they worked with marginalized communities and underrepresented groups in online advocacy efforts, ensuring that diversity of experiences within the the development of emergent tecnologies can be included. In another hand, they value partnerships and actively seek out collaborative opportunities with like-minded organizations and individuals, that’s why they are capable of bringing this project to success.

Submission Number: 187