# An EU administrative law - Speech to the European Parliament Legal Affairs Committee - Author: European Ombudsman - Date: 2023-09-15T00:00+02:00[Europe/Paris] - [URL](https://www.ombudsman.europa.eu/en/speech/en/175205) --- Chair, Honourable Members, I would like to thank the Legal Affairs committee for the invitation to speak to you today on a central topic for my office. I thank the rapporteur, honourable member Melchior, for once again highlighting an issue of considerable public interest despite the absence of significant enthusiasm for an EU administrative law from the Commission or the member states. This absence continues despite the strong endorsement for such a law from many senior legal academics, think-tanks, the European Parliament and many other professionals familiar with the complexity of EU administration. My view has not changed since the earliest days of my mandate nine years ago, when I endorsed a set of model rules drafted by a network of European academics intended to become the basis for European legislation. It is a view I set out once again in front of this committee in 2018 when it considered a very exhaustive independent study on the impact of such a law. We therefore welcome Parliament's insistence that the Commission should come forward with a law. Any new law should focus on the main principles of good administrative behaviour that promote a culture of service and of good practice. The concrete provisions of the law should allow for the fullest exercise of efficiency, of fairness and of justice. Citizens would express this in simpler terms -- that is, give me an efficient service when I come to you and if I have problems, listen to me attentively and try to sort my problems out quickly and fairly. For citizens, the law is simply an instrument to ensure these expectations are understood and met. They also understand that the most important thing is the attitude, the mindset of the officials they deal with. The officials should understand the purpose and spirit of the law and not seek to undermine it. There also needs to be flexibility and good judgement. Let me give you two examples of why this matters. The first concerns the way in which good laws have been undermined by a cultural bias that works against the original intent of a law. The EU's access to documents regulation sets out clear and reasonable time limits for responding to requests for documents, and contains clear grounds on which public access can be denied, grounds that have been tested and refined through case law. Nevertheless, my office has documented in detail the systematic delays in processing appeals to refusal decisions, often in cases of high public interest. Here, lengthy delays have become the norm, in some cases of 18 months or more. When this is the experience of journalists investigating EU decision-making or citizens concerned about how legislation has been influenced, access delayed is access denied. Frustratingly, the grounds frequently cited for denying access are based on legal reasoning already been dismissed by European Court of Justice in similar. Here we see the clear limitations of inert legislation when faced with an institutional mind-set that prioritises the control of the flow of information over the precise dictates of the law. Secondly, as this report notes, we are now at the start of a possible revolution in public administration thanks to the rapid developments in automated decision-making systems and general artificial intelligence. The use of AI in public decision -making poses immense challenges in terms of potential racial and gender bias in decision-making, and the lack of redress and accountability for routine decisions made by algorithms, particularly for the most vulnerable members of society. There are also challenges around data protection. Concerns about the use of AI and automated decision-making (ADM) in government and public services have increased, in particular in relation to the criminal justice system and 'predictive policing'; to migration control by EU border agencies and the use of controversial facial recognition, emotion recognition and biometric technologies; its use in social welfare systems to detect and prevent fraud; and to detect corruption in public administration, for example in public procurement. It is clear that there are major advantages for public administrations to improve the quality and timeliness of their services. As the European Commission recognised when it proposed the EU AI Act last year, algorithmic machine learning will transform many areas of our lives very rapidly and we need to think about the regulations and norms that will guide its development and implementation. For EU institutions, this is all in its infancy, as shown by an initiative by my office last year on the impact of AI on EU public administration, with a very limited adoption of AI and ADM technologies by the institutions and agencies. But this is likely to grow exponentially with significant consequences for accountability. How will these systems meet the requirements for human oversight, what does the guarantee of a statement of reasons for any decision mean if there is no or very poor understanding of how an algorithm made the decision? If we cannot future-proof any administrative law in this respect, then we must leave a large role for discretion and good judgement. But there also reasons to believe that the case for an EU administrative law is more compelling now than ever. The complexity and fragmentation of administration and administrative rules were demonstrated in the study commissioned by the Parliament, but this if anything has increased in ways that are not sustainable and even harmful to the objectives that the rules are designed to serve. One example concerns the ethics rules that aim to regulate the behaviour of officials. A complex, fragmented and partially overlapping system of rules for multiple institutions places a high cognitive and administrative burden on those who are subject to them. Every new system of rules raises questions around adequate training, implementation, enforcement, and resources. This can lead to "integrity fatigue" and lower compliance, especially when the anticipated benefits of these rules are not immediately evident. The Commission's proposal for an inter-institutional ethics body, is in part overdue recognition of this phenomenon, although it is still not clear whether it will add to or reduce the complexity of the system. The emphasis on harmonising standards in that proposal also downplays the importance of getting officials to internalise and model these rules and standards. A 2019 Court of Auditors report found that, despite two decades of ethics initiatives, levels of awareness and training are still low and there is an urgent need to simplify and streamline to take account of how human beings, learn and respond to incentives in their everyday working life. Finally, one of the strongest arguments for an EU administrative law is the need to bring clarity and legal certainty to question of what the lawyers call "composite procedures" or "composite administration" that is, those large parts of EU decision-making that involve cooperation between administrations or agencies at EU and national level. The pooling of EU and member states competences and funds has become the norm in large-scale funding programmes such as Next Generation EU, as well as in politically-sensitive and morally complex areas such as migration management and development funding. As we saw in our inquiry into migrant reception centres on the Greek islands earlier this year, this pooling of responsibilities between member states and the Commission can make it more difficult to disentangle the decision-making process with consequent gaps in accountability. As a "Team Europe" approach becomes the preferred method of dealing with fast-moving and complex phenomena in external relations, as shown by the memorandum of understanding concluded with the Tunisian government in June, we can expect even more confusion as to who is responsible for what. That may end up obscuring accountability precisely on those issues where it is most desperately needed. If an EU administrative law can help to address these accountability deficits, it will be most welcome addition. I wish the rapporteur all the best with this file and hope the committee will have a positive debate on the final content of this report.