As machine learning models become increasingly embedded in societal infrastructure, auditing them for bias is of growing importance. However, in real-world deployments, auditing is complicated by the fact that model owners may adaptively update their models in response to changing environments, such as financial markets. These updates can alter the underlying model class while preserving certain properties of interest, raising fundamental questions about what can be reliably audited under such shifts. In this work, we study group fairness auditing under arbitrary updates. We consider general shifts that modify the pre-audit model class while maintaining invariance of the audited property. Our goals are two-fold: (i) to characterize the information complexity of allowable updates, by identifying which strategic changes preserve the property under audit; and (ii) to efficiently estimate auditing properties, such as group fairness, using a minimal number of labeled samples. We propose a generic framework for PAC auditing based on an Empirical Property Optimization (EPO) oracle. For statistical parity, we establish distribution-free auditing bounds characterized by the SP dimension, a novel combinatorial measure that captures the complexity of admissible strategic updates. Finally, we demonstrate that our framework naturally extends to other auditing objectives, including prediction error and robust risk.

Auditability is defined as the capacity of AI systems to be independently assessed for compliance with ethical, legal, and technical standards throughout their lifecycle. The chapter explores how auditability is being formalized through emerging regulatory frameworks, such as the EU AI Act, which mandate documentation, risk assessments, and governance structures. It analyzes the diverse challenges facing AI auditability, including technical opacity, inconsistent documentation practices, lack of standardized audit tools and metrics, and conflicting principles within existing responsible AI frameworks. The discussion highlights the need for clear guidelines, harmonized international regulations, and robust socio-technical methodologies to operationalize auditability at scale. The chapter concludes by emphasizing the importance of multi-stakeholder collaboration and auditor empowerment in building an effective AI audit ecosystem. It argues that auditability must be embedded in AI development practices and governance infrastructures to ensure that AI systems are not only functional but also ethically and legally aligned.

Abstract--F ederated learning (FL) has attracted increasing attention to mitigate security and privacy challenges in traditional cloud-centric machine learning models specifically in healthcare ecosystems. FL methodologies enable the training of global models through localized policies, allowing independent operations at the edge clients' level. Conventional first-order FL approaches face several challenges in personalized model training due to heterogeneous non-independent and identically distributed (non-iid) data of each edge client. Recently, second-order FL approaches maintain the stability and consistency of non-iid datasets while improving personalized model training. This study proposes and develops a verifiable and auditable optimized second-order FL framework BFEL (blockchain enhanced federated edge learning) based on optimized FedCurv for personalized healthcare systems. FedCurv incorporates information about the importance of each parameter to each client's task (through Fisher Information Matrix) which helps to preserve client-specific knowledge and reduce model drift during aggregation. Moreover, it minimizes communication rounds required to achieve a target precision convergence for each edge client while effectively managing personalized training on non-iid and heterogeneous data. The incorporation of Ethereum-based model aggregation ensures trust, ver-ifiability, and auditability while public key encryption enhances privacy and security . Experimental results of federated CNNs and MLPs utilizing Mnist, Cifar-10, and PathMnist demonstrate the high efficiency and scalability of the proposed framework. I. Introduction Traditional machine learning (ML) methodologies necessitate training on data consolidated within a single data repository, which may be either centralized or distributed [1].

Artificial intelligence (AI) has matured as a technology, necessitating the development of responsibility frameworks that are fair, inclusive, trustworthy, safe and secure, transparent, and accountable. By establishing such frameworks, we can harness the full potential of AI while mitigating its risks, particularly in high-risk scenarios. This requires the design of responsible AI systems based on trustworthy AI technologies and ethical principles, with the aim of ensuring auditability and accountability throughout their design, development, and deployment, adhering to domain-specific regulations and standards. This paper explores the concept of a responsible AI system from a holistic perspective, which encompasses four key dimensions: 1) regulatory context; 2) trustworthy AI technology along with standardization and assessments; 3) auditability and accountability; and 4) AI governance. The aim of this paper is double. First, we analyze and understand these four dimensions and their interconnections in the form of an analysis and overview. Second, the final goal of the paper is to propose a roadmap in the design of responsible AI systems, ensuring that they can gain society's trust. To achieve this trustworthiness, this paper also fosters interdisciplinary discussions on the ethical, legal, social, economic, and cultural aspects of AI from a global governance perspective. Last but not least, we also reflect on the current state and those aspects that need to be developed in the near future, as ten lessons learned.

Audits contribute to the trustworthiness of Learning Analytics (LA) systems that integrate Artificial Intelligence (AI) and may be legally required in the future. We argue that the efficacy of an audit depends on the auditability of the audited system. Therefore, systems need to be designed with auditability in mind. We present a framework for assessing the auditability of AI-integrating systems that consists of three parts: (1) Verifiable claims about the validity, utility and ethics of the system, (2) Evidence on subjects (data, models or the system) in different types (documentation, raw sources and logs) to back or refute claims, (3) Evidence must be accessible to auditors via technical means (APIs, monitoring tools, explainable AI, etc.). We apply the framework to assess the auditability of Moodle's dropout prediction system and a prototype AI-based LA. We find that Moodle's auditability is limited by incomplete documentation, insufficient monitoring capabilities and a lack of available test data. The framework supports assessing the auditability of AI-based LA systems in use and improves the design of auditable systems and thus of audits.

Visa is a world leader in digital payments, facilitating more than 215 billion payments transactions between consumers, merchants, financial institutions and government entities across more than 200 countries and territories each year. Our mission is to connect the world through the most innovative, convenient, reliable and secure payments network, enabling individuals, businesses and economies to thrive. When you join Visa, you join a culture of purpose and belonging – where your growth is priority, your identity is embraced, and the work you do matters. We believe that economies that include everyone everywhere, uplift everyone everywhere. Your work will have a direct impact on billions of people around the world – helping unlock financial access to enable the future of money movement.

The past few years have brought much hand wringing and arm waving about artificial intelligence (AI), as business people and technologists alike worry about the outsize decisioning power they believe these systems to have. As a data scientist, I am accustomed to being the voice of reason about the possibilities and limitations of AI. In this article I'll explain how companies can use blockchain technology for model development governance, a breakthrough to better understand AI, make the model development process auditable, and identify and assign accountability for AI decisioning. While there is widespread awareness about the need to govern AI, the discussion about how to do so is often nebulous, such as in "How to Build Accountability into Your AI" in Harvard Business Review: A healthy ecosystem for managing AI must include governance processes and structures.... Accountability for AI means looking for solid evidence of governance at the organizational level, including clear goals and objectives for the AI system; well-defined roles, responsibilities, and lines of authority; a multidisciplinary workforce capable of managing AI systems; a broad set of stakeholders; and risk-management processes. Additionally, it is vital to look for system-level governance elements, such as documented technical specifications of the particular AI system, compliance, and stakeholder access to system design and operation information.

The research and development (R&D) phase of building an AI model to address a business problem is characterized by rapid exploration and iteration. Everything is on the table and experimentation is encouraged, from understanding how to frame the problem, to determining how to most effectively use the data on hand, to discovering the model architecture with the best performance. In stark contrast to this, the operationalization phase of AI model development requires that the model be completely characterized, produce reproducible results, and be stable for integration in automation processes. Model versioning best practices and version control tools are essential to successfully navigating and overcoming this gap between R&D and production engineering. The practice of version control is nothing new.

In recent years, the banking industry has been at the forefront of AI and ML adoption. According to an Economist Intelligence Unit adoption study, 54% of banks and financial institutions with more than 5,000 employees have adopted AI. But AI and ML adoption has not been easy. Difficulty in deployment has been exacerbated by the growing number of new AI platforms, languages, frameworks, and hybrid compute infrastructure. Add to this the fact that models are being developed by staff in multiple business units and AI teams, making it difficult to ensure that the proper risk and regulatory controls and processes are enforced.

Group fairness metrics can detect when a deep learning model behaves differently for advantaged and disadvantaged groups, but even models that score well on these metrics can make blatantly unfair predictions. We present smooth prediction sensitivity, an efficiently computed measure of individual fairness for deep learning models that is inspired by ideas from interpretability in deep learning. smooth prediction sensitivity allows individual predictions to be audited for fairness. We present preliminary experimental results suggesting that smooth prediction sensitivity can help distinguish between fair and unfair predictions, and that it may be helpful in detecting blatantly unfair predictions from "group-fair" models.