D&I in AI – Industry Collaboration

Operationalizing the D&I in AI Guidelines in Co-Design Workshops

Non-adherence to practices of diversity and inclusion in the building, deployment, and use of AI systems have been shown to cause digital redlining, discrimination, and algorithmic oppression, leading to AI systems being perceived as untrustworthy and unfair. Whether it’s the recognition technologies that exhibit racist behaviour or manifest sexist outcomes, ignoring the social context of the AI use and neglecting the inclusion of humans with diverse attributes during design and development can also create problems. A review of 47 ethical guidelines that have been developed by governments, inter-governments, the private sector, civil society and multi-stakeholders, has shown that in its current state, AI ethics guidelines are focusing disproportionately on issues of algorithmic decision-making, while the fairness, accountability, sustainability, and transparency of the business decision-making contexts in which AI systems are situated remain seriously undermined by competitive and speculative norms, ethics washing, corporate secrecy, and other harmful business practices. Diversity is mostly considered as a part of showing the intention of fairness, however, we argue that achieving “diversity and inclusion” in the AI ecosystem must be placed as the core intention.

Diversity and Inclusion in AI can no longer be thought of as a “nice to have” feature, rather a “must have” principle that all AI technology vendors and developers should aspire to gain a deeper understanding of, and put into practice, in their innovations and transformation initiatives.

Simply having guidelines is not enough; they must be actively integrated into the development and deployment of AI systems to ensure ethical, fair, and representative outcomes. By operationalizing D&I, we can move beyond theoretical concepts and translate them into tangible actions and results. This ensures that AI systems are developed with a deep understanding of diverse perspectives, leading to more innovative, effective, and equitable solutions that cater to the needs of a diverse society.

Our team is actively engaged with industry partners through co-design workshops to tailor D&I in AI guidelines for their specific contexts. By collaborating closely with partners, we ensure that the D&I guidelines are not only relevant but also practical and actionable within their unique environments. This hands-on approach facilitates the integration of D&I principles into the fabric of AI development, enabling our industry partners to create more inclusive and equitable AI systems. Through these workshops, we foster a shared understanding and commitment to operationalizing D&I in AI, driving meaningful change across various sectors.

Industry Collaborations

Open allClose all

D&I in AI for Recruitment

We have successfully conducted a co-design workshop with a large international recruitment industry SEEK to tailor the D&I in AI guidelines for their specific use cases. Diversity and Inclusion (D&I) in AI for recruitment is critical to ensure that hiring processes are fair, unbiased, and representative of the diverse talent pool. AI-driven recruitment tools, if not carefully designed with D&I principles, can perpetuate existing biases, leading to discriminatory practices. Contextualized guidelines developed through co-design workshops are a way forward, as they involve stakeholders from various backgrounds in the creation process, ensuring that the guidelines are tailored to address the specific needs and challenges of different organizations. This collaborative approach ensures that AI recruitment tools are developed with a deep understanding of diversity and inclusion, promoting equity and fairness in hiring practices.

D&I in AI for Media

We have successfully established collaboration and will soon conduct a co-design workshop with a leading media organization in Australia ABC to tailor the Diversity and Inclusion (D&I) in AI guidelines for their specific needs in content creation and distribution. D&I in AI for media is crucial to ensure that content is inclusive, unbiased, and representative of diverse audiences. AI-driven tools used in media, if not carefully designed with D&I principles, can perpetuate stereotypes and exclude underrepresented groups. Contextualized guidelines developed through co-design workshops are essential, as they involve stakeholders from various backgrounds in the creation process, ensuring that the guidelines are tailored to address the specific needs and challenges of the media industry.

D&I in AI for Finance

We have started a collaboration with one of the top insurance companies in Australia. The critical importance of diversity and inclusion in AI for insurance cannot be overstated. AI-driven insurance models and algorithms, if not designed with diversity and inclusion principles, can inadvertently perpetuate biases and lead to various forms of discrimination. This can result in unfair pricing, coverage disparities, and exclusion of certain groups based on race, gender, or other protected attributes. By ensuring diversity and inclusion in AI development, insurance companies can create more equitable and inclusive products that accurately reflect and serve the diverse needs of their customers. Ignoring these principles not only undermines ethical standards but also exposes insurers to legal and reputational risks associated with discriminatory practices.

D&I in AI for Sports

We participated in a workshop on 1st March 2024, for the drafting of a position paper for Responsible AI in Sports, organised by the Australian Sports Commission along with Data61. AI in sports has the potential to revolutionize the industry by optimizing performance, preventing injuries, identifying talent, engaging fans, and analyzing games. However, it also presents ethical challenges, including issues of bias, fairness, privacy, transparency, and accountability. To address these challenges, Diversity and Inclusion (D&I) principles should be central to the ethical framework guiding the use of AI in sports. Currently, there is a significant research gap in integrating the ‘ethics of AI’ into sports, indicating a need for further exploration and development in this area to ensure that AI is used responsibly and equitably in the sports sector.

Media Report by Australian Sports Commission: ASC workshop shines light on responsible AI in sport | Australian Sports Commission (sportaus.gov.au)

Microsoft AI Copilot Evaluation

Dr Muneera Bano is currently leading the research project on the evaluation of Microsoft AI copilot. Microsoft Copilot is designed to boost productivity and creativity across various applications. The Australian Government is conducting a six-month trial of Microsoft 365 Copilot, making it one of the first governments in the world to deploy AI services. At CSIRO, Bano and her team are leading the research study on the six-month trial in 2024 to evaluate Microsoft Copilot’s impact on productivity, efficiency, effectiveness, ethical aspects, and user satisfaction. This trial will engage almost 300 staff members, providing a diverse sample for analysis.

Data61-Google Event on Diversity and Inclusion

As one of the DIB officers for Data61, Dr Muneera Bano is leading a project in collaboration with Google Australia to bring together university representatives in computing and AI fields, to elicit the best practices of D&I.

Event Details:

Date: 1st May 2024
Time: 10:00 AM – 4:00 PM
Venue: Google Office, Sydney
Theme: Fostering Diversity, Inclusion, and Belongingness in Computing Domains in Academia

Objective:

The upcoming event aims to address the urgent need for enhanced diversity and inclusion in computing fields. This initiative is in response to the findings from the “Australian Universities Accord Report” and the “Pathways to Diversity in STEM“. The focus will be on creating a collaborative space for university representatives to exchange experiences, challenges, and successes in promoting diversity and inclusion in academia.

Agenda:

The event will feature a series of roundtable conversations, breakout sessions, and talks covering topics such as:

Creating inclusive policies
Designing curricula that promote diversity
Fostering supportive environments for staff and students

Key Highlights:

Australian Universities Accord Report: The report underscores the importance of an inclusive environment for attracting and retaining a diverse workforce in computing fields. It advocates for the implementation of diversity and inclusion strategies to empower underrepresented groups, such as women, Indigenous peoples, and people from diverse backgrounds.
Pathways to Diversity in STEM: This report provides recommendations and highlights the strategic significance of diversity and inclusion in computing. It emphasizes the role of inclusive practices in promoting innovation, improving problem-solving skills, and preventing biases in algorithms and systems.