The European Commission unveiled its proposal last April for the regulation of Artificial Intelligence applications in the EU, introducing a number of important safeguards for a balanced approach to AI. However, there are still important concerns about the quality of data that is used for training AI in making observations that can inform policy. Data that contains inaccuracies could allow room for cognitive bias and undermine effective evidence-based policymaking.
Addressing these gaps will be key to ensuring that researchers, and other actors such as civil society, can cross-monitor datasets for bias and misinformation. Otherwise, replicating bias in our world into the world of AI could be catastrophic, and an obstacle to a smooth digital transition for the EU.
This roundtable discussion will discuss the key challenges that researchers and other stakeholders identify in their work with regard to accessing important data and monitoring it for bias and misinformation. The event will discuss how more transparency in AI could give way to data and monitoring for bias and misinformation, and thus contribute to more transparent and ethical policy advice and policies.
This workshop is part of the PRO-RES final event series to discuss the findings of the PRO-RES Project in different policy settings, including Artificial Intelligence. PRO-RES encourages policymakers and advisors to seek evidence for their decisions from research that has been conducted ethically, responsibly, and with integrity.