Introduction

The report has been published by FRA-European Union Agency for Fundamental Rights in December 2020 and focuses on safeguarding fundamental rights. It stems from FRA’s project on Artificial intelligence, big data and fundamental rights.

At a time when AI is being extensively used, even amidst the COVID-19 health crisis, questions are continuously being raised regarding the challenges and threats that this new technology is posing to human and fundamental rights. Based on use cases of AI in selected areas, the report analyses the implications on fundamental rights when using the technology. It also takes into account 91 interviews conducted with officials, both in the private and public sector, in selected EU member states. Based on these, the report draws certain conclusions and opinions regarding fundamental rights with respect to adoption of AI.

In recent years, the use of AI systems infringing on privacy and data protection has gained much attention. However, in the process of securing privacy and data protection what often gets lost in the process is a range of rights in areas such as equality, non-discrimination, human dignity and social security. Keeping this in mind, the FRA is of the opinion that EU legislators should respect the full spectrum of fundamental rights as laid out in the Charter and EU treaties while adopting new legislations on AI. Legal safeguards need to be provided in case of arbitrary interference with fundamental rights. The FRA is also of the opinion that the EU legislators should make impact assessment mandatory for both private and public sectors that are using services of AI. Such assessments should be based on established good practices followed in other fields or sectors.

The other opinions highlighted in the report are putting in place an effective accountability system to monitor and address any negative impact of AI systems on fundamental rights, encouraging public administration and companies to assess any discriminatory outcomes in the process of using AI systems and funding such research that helps in identifying such outcomes, ensuring effective access to justice for individuals in case of dispute with AI-backed decisions and providing guidance and support to effectively implement GDPR provisions to the use of AI.

The report also gives an in-depth analysis of the use-cases and is of the opinion that the single most reason for use of AI is increase in efficiency coupled with minimization of human errors. One of the interesting use-case is in the area of predictive policing where the aim is to anticipate crime in advance. Under this, the FRA strongly iterates the need for preventing unlawful profiling and use of sensitive data (such as facial recognition) with caution so as to prevent possible misuse. 

Relevance of the Report

The report gives an excellent overview of the use of AI in the EU across a range of sectors such as healthcare, social benefits, predictive policing and targeted advertising. It also provides adequate information on the purpose and practical application of AI technologies, the awareness of fundamental rights issues and potential safeguards when it comes of use of AI and the future plans with respect to the adoption of the technology. Above all, in a very clear yet crisp manner, it delivers a fundamental rights-based analysis of case studies carried out in five EU member states (Finland, France, Spain, Estonia and the Netherlands). This analysis will help other member states of EU to assess the compatibility of fundamental rights and chart out the required actions to guarantee fundamental rights. 

Key Takeaways

· The widespread use of AI does, at times, infringe upon fundamental rights.

· There is a need for flexible impact assessments of AI related technologies that can adapt as per the situation. In this regard, it can be noted that there are several practical tools like checklists, list of questions, risk management frameworks etc. to assess the impact.

· Accountability of AI is important which means organizations using AI should be able to explain the basis on which AI-driven decision is taken. In other words, it is extending the right to know for the common public.

· Spread of awareness of potential rights implications apart from data privacy and non-discrimination. Areas such as access to justice, human dignity and consumer protection can also be at risk with use of AI related technologies.