On 21 April 2021, the European Commission proposed the first legal framework on Artificial Intelligence (AI) to address the risks posed by this emerging method of computation. The Commission proposed a Regulation known as the AI Act. The proposed AI Act considers not only machine learning, but expert systems and statistical models long in place. Under the proposed AI Act, new obligations are set to ensure transparency, lawfulness, and fairness. Their goal is to establish mechanisms to ensure quality at launch and throughout the whole life cycle of AI-based systems, thus ensuring legal certainty that encourages innovation and investments on AI systems while preserving fundamental rights and values. A standardisation process is ongoing: several entities (e.g., ISO) and scholars are discussing how to design systems that are compliant with the forthcoming Act, and explainability metrics play a significant role. Specifically, the AI Act sets some new minimum requirements of explicability (transparency and explainability) for a list of AI systems labelled as “high-risk” listed in Annex III. These requirements include a plethora of technical explanations capable of covering the right amount of information, in a meaningful way. This paper aims to investigate how such technical explanations can be deemed to meet the minimum requirements set by the law and expected by society. To answer this question, with this paper we propose an analysis of the AI Act, aiming to understand (1) what specific explicability obligations are set and who shall comply with them and (2) whether any metric for measuring the degree of compliance of such explanatory documentation could be designed. Moreover, by envisaging the legal (or ethical) requirements that such a metric should possess, we discuss how to implement them in a practical way. More precisely, drawing inspiration from recent advancements in the theory of explanations, our analysis proposes that metrics to measure the kind of explainability endorsed by the proposed AI Act shall be risk-focused, model-agnostic, goal-aware, intelligible, and accessible. Therefore, we discuss the extent to which these requirements are met by the metrics currently under discussion.

In this paper we introduce a new class of software tools engaged in delivering successful explanations of complex processes on top of basic Explainable AI (XAI) software systems. These tools, that we call cumulatively Explanatory AI (YAI) systems, enhance the quality of the basic output of a XAI by adopting a user-centred approach to explanation that can cater to the individual needs of the explainees with measurable improvements in usability. Our approach is based on Achinstein’s theory of explanations, where explaining is an illocutionary (i.e., broad yet pertinent and deliberate) act of pragmatically answering a question. Accordingly, user-centrality enters in the equation by considering that the overall amount of information generated by answering all questions can rapidly become overwhelming and that individual users may perceive the need to explore just a few of them. In this paper, we give the theoretical foundations of YAI, formally defining a user-centred explanatory tool and the space of all possible explanations, or explanatory space, generated by it. To this end, we frame the explanatory space as an hypergraph of knowledge and we identify a set of heuristics and properties that can help approximating a decomposition of it into a tree-like representation for efficient and user-centred explanation retrieval. Finally, we provide some old and new empirical results to support our theory, showing that explanations are more than textual or visual presentations of the sole information provided by a XAI.

The European AI Act has introduced specific technical documentation requirements for AI systems. Compliance with them is challenging due to the need for advanced knowledge of both legal and technical aspects, which is rare among software developers and legal professionals. Consequently, small and medium-sized enterprises may face high costs in meeting these requirements. In this study, we explore how contemporary AI technologies, including ChatGPT and an existing compliance tool (DoXpert), can aid software developers in creating technical documentation that complies with the AI Act. We specifically demonstrate how these AI tools can identify gaps in existing documentation according to the provisions of the AI Act. Using open-source high-risk AI systems as case studies, we collaborated with legal experts to evaluate how closely tool-generated assessments align with expert opinions. Findings show partial alignment, important issues with ChatGPT (3.5 and 4), and a moderate (and statistically significant) correlation between DoXpert and expert judgments, according to the Rank Biserial Correlation analysis. Nonetheless, these findings underscore the potential of AI to combine with human analysis and alleviate the compliance burden, supporting the broader goal of fostering responsible and transparent AI development under emerging regulatory frameworks.

Human environments are often regulated by explicit and complex rulesets. Integrating Reinforcement Learning (RL) agents into such environments motivates the development of learning mechanisms that perform well in rule-dense and exception-ridden environments such as autonomous driving on regulated roads. In this letter, we propose a method for organising experience by means of partitioning the experience buffer into clusters labelled on a per-explanation basis. We present discrete and continuous navigation environments compatible with modular rulesets and 9 learning tasks. For environments with explainable rulesets, we convert rule-based explanations into case-based explanations by allocating state-transitions into clusters labelled with explanations. This allows us to sample experiences in a curricular and task-oriented manner, focusing on the rarity, importance, and meaning of events. We label this concept Explanation-Awareness (XA). We perform XA experience replay (XAER) with intra and inter-cluster prioritisation, and introduce XA-compatible versions of DQN, TD3, and SAC. Performance is consistently superior with XA versions of those algorithms, compared to traditional Prioritised Experience Replay baselines, indicating that explanation engineering can be used in lieu of reward engineering for environments with explainable features.