Framework The issues related to AI are increasingly present on political and regulatory agendas, so much so that the EU is striving to become the first regional actor to regulate AI with a specific legal instrument. Its Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (‘the AI Act’), presented on 21 April 2021, will certainly be a benchmark for other jurisdictions in the development of AI. It Is therefore essential that, in addition to legal regulation, documents, such as guidelines or directives that explore and lift the veil not only on the benefits of AI, but also on its negative consequences, especially when t
Insight
EU: Unpacking the FRA's report on bias in algorithms, AI, and discrimination
February 15, 2023
Summary
The European Union Agency for Fundamental Rights (FRA) released a report on the risks of bias in AI algorithms, particularly in predictive policing and hate speech detection, and its impact on fundamental human rights. The report emphasizes the need for thorough bias assessments and human oversight to prevent discrimination and fundamental rights violations. It discusses the challenges of mitigating bias, including the lack of transparency, biased data, and feedback loops that exacerbate discrimination. The FRA concludes that in some cases, it may be impossible to eliminate bias, and algorithms should not be used without meaningful human intervention.