Research and best practices in addressing risks in AI systems have significantly progressed over the last years. This panel looks at the most challenging problems and advances in research to support fairness, accountability, transparency and equity in AI. The panel will also examine whether the AI Act’s requirements for trustworthiness will be flexible enough to address these objectives, nuanced enough to tackle the diversity of AI systems and their specific risks as well as the pace of innovation.
• Are the requirements able to tackle the socio-technical challenges of AI systems?
• What are the criteria against which the AI Act requirements will be measured?
• Are outcome-based goals an alternative?
• How can AI deployers be supported in their fairness work in practice?
• How can stakeholders impacted by AI participate in designing fairer and more responsible AI?