Combating AI bias in the labour market
The EU-funded projects AEQUITAS and BIAS are joining forces to combat AI system biases in the labour market. By working together, they aim to deepen their understanding of AI technologies’ impact on workers and inform policies accordingly. This partnership will play an important role in furthering efforts to ensure fair and transparent AI systems in Europe. AEQUITAS, which was launched in November 2022, is aiding the development of a European ecosystem that promotes fairness, transparency, equality and accountability in AI systems. The 3-year project is making this happen by providing AI developers and users with an open platform to create controlled experiments for detecting, mitigating and repairing potential issues of unfairness in AI tools. Launched at the same time as AEQUITAS, the 4-year BIAS project is empowering AI and human resource management (HRM) communities by tackling and reducing algorithmic biases. The project has three main objectives: develop reliable tools for identifying and mitigating bias in AI systems; encourage stakeholder engagement in the co-creation of technology to mitigate biases and unfairness in decision-making; and empower the AI and HRM communities through awareness raising and capacity building initiatives. Eduard Fosch-Villaronga of BIAS project partner Leiden University, the Netherlands, comments on the partnership in a news item posted on the university’s website: “By combining our efforts and expertise, we can advance the field of AI ethics and ensure that AI technologies are developed and deployed in a manner that respects the rights and dignity of all individuals in the workforce.” Researchers Gustavo Marrero and Ángel Marrero of AEQUITAS project partner Universidad de la Laguna, Spain, further discuss the benefits of collaboration: “Our collaboration with Leiden University’s eLaw Center will enable us to pool our resources and insights to develop effective strategies for mitigating biases in AI systems, ultimately fostering a more equitable and inclusive labour market.”
Sharing knowledge and experience
The two projects will also be holding separate workshops on bias in AI systems. On 24 May 2024, BIAS is hosting a workshop titled ‘Fair Futures at Work: Co-Creation and AI-driven Solutions in Governing the Hiring Process’. It will be held at the Computers, Privacy and Data Protection conference (22-24 May) in Brussels, Belgium, and will focus on the potential and limitations of technical solutions to promote fairness and trustworthiness in AI-powered hiring. The BIAS (Mitigating Diversity Biases of AI in the Labor Market) project is also organising a second workshop between 28 and 31 May, this time in Hamamatsu, Japan. Titled ‘The ambitions and limitations of AI-driven recruitment and selection: Unfolding fairness and diversity bias’, the workshop aims to explore how AI technologies shape notions of fairness and diversity bias, and seeks paths to a more equitable and just landscape. Later in the year, on 19 October, AEQUITAS (ASSESSMENT AND ENGINEERING OF EQUITABLE, UNBIASED, IMPARTIAL AND TRUSTWORTHY AI SYSTEMS) will be holding the second edition of its ‘Workshop on Fairness and Bias in AI’ at the 27th European Conference on Artificial Intelligence (19-24 October) being held in Santiago de Compostela, Spain. A call for papers is currently underway, and submissions are welcome until 15 May. For more information, please see: AEQUITAS project website BIAS project website
Keywords
AEQUITAS, BIAS, AI, diversity bias, fairness, worker, labour market