The EU AI Act Newsletter #5
29/03/22-12/04/22
Welcome to the EU AI Act Newsletter, a brief biweekly newsletter by the Future of Life Institute providing you with up-to-date developments and analyses of the proposed EU artificial intelligence law.
Legislative Process
According to EURACTIV, the two European Parliament co-rapporteurs Dragoș Tudorache and Brando Benifei finalised the AI Act draft report on 11 April. They agreed to keep the definition of AI broad, added predictive policing to prohibited practices, and included more substantial obligations and transparency requirements for public authorities when using high-risk applications. The draft report will be discussed in the two parliamentary committees on 11 May.
The French Presidency of the European Council shared several compromises on the AI Act on 4 April. EURACTIV briefly summarised the text. Some of the main things highlighted in the document are related to the use of biometric identification systems, changes in the conformity assessment procedure, exception to transparency obligations, making some information in the EU database for high-risk AI systems not public, and creating some exceptions to market monitoring and surveillance. The Presidency's proposed changes are available here by Statewatch.
The Committee on Transport and Tourism (TRAN) at the European Parliament published their draft opinion on the AI Act. The committee's primary suggestions are to ensure that the AI Act does not overlap with sectoral legislation by imposing double/conflicting obligations on transport actors; promoting the development of, and upholding, international standards, that are
particularly important for the transport sector; and fostering research and innovation to ensure the EU’s transport sector develops its own know-how in the implementation of AI, while upholding the highest ethical standards.
Several changes have been proposed to MEP Eva Maydell's (Committee on Industry, Research and Energy) draft opinion report on the AI Act by other members of the European Parliament. These amendments are available here and here. A lot of changes have also been suggested to MEP Axel Voss’ (Committee on Legal Affairs) draft opinion report. These can be found here, here and here.
Analyses
Susana Solís Pérez (ES, RE), rapporteur for the opinion of the Committee on Environment, Public Health and Food Safety (ENVI) on the AI Act, published an op-ed in The Parliament Magazine. She says that AI can be pivotal in the fight against climate change. According to her, AI will lead the way in reaching climate neutrality through monitoring the environment, designing smarter and greener cities, reducing energy consumption, developing more efficient transportation and making waste management more intelligent. She demands that the protection of the environment be considered a high-risk factor in the AI Act.
Science|Business summarised a recent Data Rules workshop where experts discussed issues related to the global governance of AI and the effect of the AI Act on that. Brando Benifei, one of the European Parliament’s rapporteurs for the Act, stated that it is important to avoid a situation where there are two different contexts of application, standards and regulation exist simultaneously. Vidushi Marda, senior programme officer at ARTICLE 19, does not think that India will see the AI Act as a blueprint, because India is focused on safety, security and business rather than on human rights. Alex Engler, a fellow at the Brookings Institution, thinks that the AI Act might have a limited global impact.
Ada Lovelace Institute, UK-based research institute working in data, policy and regulation, published several policy papers on the AI Act. In the first, they critique the approach of the AI Act which would work well for tangible products but is questionable for AI. They argue that AI is not a product but a service that learns and changes, AI systems can be general purpose meaning that they can be applied to different contexts, and AI is part of a larger system or platform. The second paper provides specific recommendations for EU policymakers for the final version of the Act.
Verfassungsblog shared an op-ed that discusses the issue of AI enabling the surveillance and manipulation of humans at a scale never possible before. The op-ed refers to several regulations proposed by the European Commission to deal with this issue: the Digital Service Act (DSA), the Regulation on the Transparency and Targeting of Political Advertising (TTPA) and the Regulation on Artificial Intelligence (AI Act). They argue that the AI Act presents a grand opportunity to address AI-driven manipulation.