GenCon
Responsible Research beyond AI Technology
The strategic project “Gender and Diversity Considerations for Responsible Research beyond AI Technology”, in short “GenCon” was intended to raise awareness among researchers about how gender and diversity perspective need to be considered in application-oriented (AI) research, especially those that use artificial intelligence (AI) technology.
For decades, the European Commission has pursued a three-pronged strategy to achieve gender equality in science and research (see also Schiebinger et al. 2011):
"Fix the Numbers": Creating balanced teams by increasing the participation of previously underrepresented genders (e.g., more women in technical professions).
"Fix the Institutions": Structural change in research organizations through the implementation of gender equality plans.
"Fix the Knowledge": Integrating gender (and "sex," including biological sex) as a dimension in research content.
GenCon specifically addressed the third – and often under-researched – aspect and initially analyzed the potential of gender and diversity in Pro2Future research. These analysis results led, in a second step, to a joint knowledge production process (“co-creation of gender knowledge,” Dahmen-Adkins & Thaler 2022) to sensitize Pro2Future researchers to the topic and to increasingly consider and integrate gender and diversity in future research projects. The theoretical framework of gender knowledge (Wetterer 2009) helped to specifically locate the existing body of knowledge within the potential analysis and to build the gender and diversity workshop on it. Through the project several valuable resources available at the center as well as six fields of intervention and three strategic fields of action were assessed respectively defined. The identified three strategic fields of action provide the basis for future measures.
Goals
The strategic project of Pro2Future and IFZ aimed to raise awareness among Pro2Future researchers about how gender and diversity can be considered and integrated into application-oriented research projects, especially those using artificial intelligence (AI) technology.
The project was designed as a research project that, through collected data, provides an evidence-based foundation for embedding gender and diversity in Pro²Future's research areas and projects. GenCon was a research project in the field of "Science, Technology, and Society Studies" (STS for short, see, e.g., Bammé 2009) and is based on the insight that technological developments are socially influenced and, conversely, technologies influence societal and social processes (Harald Rohracher speaks of "The mutual shaping of design and use," 2006; Langdon Winner's famous essay asks "Do artifacts have politics?", 1980; Bruno Latour argues that technologies are "society made durable," cited in Bammè 2009, p. 132). An important area of STS for GenCon is gender in science and research, addressing, among other things, how gender relations are inscribed in technological innovations ("gender scripts") and thus how power relations can be reproduced (see, e.g., Wajcman 2004, Zorn et al. 2007).
Approach
GenCon uses empirical data to enable an evidence-based analysis of potential for the possible integration of gender and diversity, as well as "Responsible Research and Innovation" (Föger et al. 2016), into Pro2Future's research. Continuous feedback loops with the Gender and Diversity Working Group were particularly important. This GenCon working group consisted of the scientific and commercial management of Pro2Future, as well as the heads of the Human Resources department and the Office Management.
Expected and Achieved Results
The potential analysis from project concepts, descriptions, and interviews, as well as the results of the joint knowledge production can be translated into three strategic fields of action:
Potential 1: (Social) Scientific Methods
The technically correct application of social scientific methods can both help with the involvement of users and stakeholders throughout the research process and improve scientific quality through the objectivity of the processes (Steinke 2003). This also includes stakeholder mapping, which can be used to methodically identify and engage diverse, relevant stakeholders in a target-group-specific manner (see Anslinger et al. 2022). Balanced stakeholder involvement can, on the one hand, minimize biases in AI and serve as the basis for responsible research and, in particular, trustworthy AI research (AI HLEG 2019, Anslinger et al. 2022).
Potential 2: Responsible Research and Innovation
The research topics of P2F offer many points of contact with the concept of Responsible Research and Innovation (Föger et al. 2016). Consistent stakeholder engagement and a human-centered approach, as well as the pursuit of socially relevant topics such as gender equality and sustainability, could be logical starting points for P2F. The consideration of ethical standards and criteria offers enormous potential, particularly in the field of AI (“Fair AI,” Anslinger 2021; “AI Ethics,” Funk 2021; “Trustworthy AI,” AI-HLEG 2019; “AI ACT,” Council of the EU 2023).
Potential 3: Diversity-Sensitive Human-Centered Design
The analysis of the usage contexts of real users and affected stakeholders, i.e., "human-centered design" (Hartson et al. 2012, Rohracher 2005; specifically, human-computer interaction, e.g., Seinfeld et al. 2021), contrasts with "executive-centered AI" (Henriksen & Blond 2023), which overemphasizes the management perspective in companies. The transdisciplinary inclusion of actual user and stakeholder needs, beyond stereotypes, generates "socially robust knowledge" (Nowotny 1999), leads to more robust technologies, and helps improve the acceptance of these technologies.
