Artificial intelligence (AI) and algorithmic decision making are having a profound impact on our daily lives. These systems are vastly used in different high-stakes applications like healthcare, business, government, education, and justice, moving us toward a more algorithmic society. However, despite so many advantages of these systems, they sometimes directly or indirectly cause harm to the users and society. Therefore, it has become essential to make these systems safe, reliable, and trustworthy. Several requirements, such as fairness, explainability, accountability, reliability, and acceptance, have been proposed in this direction to make these systems trustworthy. This survey analyzes all of these different requirements through the lens of the literature. It provides an overview of different approaches that can help mitigate AI risks and increase trust and acceptance of the systems by utilizing the users and society. It also discusses existing strategies for validating and verifying these systems and the current standardization efforts for trustworthy AI. Finally, we present a holistic view of the recent advancements in trustworthy AI to help the interested researchers grasp the crucial facets of the topic efficiently and offer possible future research directions.

Trustworthy Artificial Intelligence: A Review

An empirical study of the rise of big data in business scholarship

In this article, we examine the concept of non-personal data from a law and computer science perspective. The delineation between personal data and non-personal data is of paramount importance to determine the GDPR’s scope of application. This exercise is, however, fraught with difficulty, also when it comes to de-personalised data – that is to say data that once was personal data but has been manipulated with the goal of turning it into anonymous data. This article charts that the legal definition of anonymous data is subject to uncertainty. Indeed, the definitions adopted in the GDPR, by the Article 29 Working Party and by national supervisory authorities diverge significantly. Whereas the GDPR admits that there can be a remaining risk of identification even in relation to anonymous data, others have insisted that no such risk is acceptable. A review of the technical underpinnings of anonymisation that is subsequently applied to two concrete case studies involving personal data used on blockchains, we conclude that there always remains a residual risk when anonymisation is used. The concluding section links this conclusion more generally to the notion of risk in the GDPR.

They who must not be identified—distinguishing personal from non-personal data under the GDPR

Value sensitive design (VSD) is an established method for integrating values into technical design. It has been applied to different technologies and, more recently, to artificial intelligence (AI). We argue that AI poses a number of challenges specific to VSD that require a somewhat modified VSD approach. Machine learning (ML), in particular, poses two challenges. First, humans may not understand how an AI system learns certain things. This requires paying attention to values such as transparency, explicability, and accountability. Second, ML may lead to AI systems adapting in ways that ‘disembody’ the values embedded in them. To address this, we propose a threefold modified VSD approach: (1) integrating a known set of VSD principles (AI4SG) as design norms from which more specific design requirements can be derived; (2) distinguishing between values that are promoted and respected by the design to ensure outcomes that not only do no harm but also contribute to good, and (3) extending the VSD process to encompass the whole life cycle of an AI technology to monitor unintended value consequences and redesign as needed. We illustrate our VSD for AI approach with an example use case of a SARS-CoV-2 contact tracing app.

/pdf/mapping-value-sensitive-design-onto-ai-for-social-good-13jemhe7zo.pdf

Mapping value sensitive design onto AI for social good principles

Trust and Crime in Information Societies

Introduction This era of big data analytics promises many things. In particular, it offers opportunities to extract hidden value from unstructured raw datasets through novel reuse. The reuse of personal data is, however, a key concern for data protection law as it involves processing for purposes beyond those that justified its original collection, at odds with the principle of purpose limitation. The issue becomes one of balancing the private interests of individuals and realizing the promise of big data. One way to resolve this issue is to transform the personal data that will be shared for further processing, such as data mining, into ―anonymous information‖ to use an EU legal term. "Anonymous information" is outside the scope of data protection laws in the EU, and is also carved out from privacy laws in many other jurisdictions worldwide. The foregoing solution works well in theory, but only as long as the output potential from the data still retains utility, which is not necessarily the case in practice. This is because the value or knowledge that can be gained from analyzing datasets (particularly using automatic algorithmic software) is maximized by virtue of finding patterns, basically linking relationships between data points. Anonymization, by contrast, aims to delink such data point relationships where they relate to informational knowledge that can be gleaned in respect of specific persons and their identities. This leaves those in charge of processing the data with a problem: how can they ensure that anonymization is conducted effectively on the data on their possession, while retaining that data‘s utility for potential future disclosure to, and further processing by, a third party?

Anonymous data v. personal data—a false debate: an EU perspective on anonymization, pseudonymization and personal data

Identity crisis: Global challenges of identity protection in a networked world

Means of identification are growing rapidly with new digital and online tracking capabilities, and other emerging technologies of personal verification and authentication. Today, we emit a wide spectrum of direct - but also indirect – identifiers which, in recorded form, either alone or in combinations, can lead to us being known in different ways. Indeed, in a world that is becoming increasingly hyper-connected and digitally-surveyed, internet-connected devices leak a constant stream of data with which individual users (and proximate others in the data-collection vicinities) can be associated. Furthermore, the more data about a person that are collected, the easier it can become for further information to be inferred about them, linked to them, as well as for them to be singled out from others. Along with the development of data analytical techniques, this has led to recognition that scientific and technological advances are making it increasingly difficult to de-identify personal data and guarantee that the individuals to whom such data relates may not be later re-identified from it. The legal, privacy, and regulatory challenges that flow from these facts have led to the present research. They point to the fact that there is confusion about the legal definition of personal data reliant upon the concept of identification capabilities from information, in interpretation and practical application (in determining what data comes within its scope and data protection obligations apply), when confronted with new technological realities. More specifically, such challenges point to the critical importance of reconsidering the requirement of being identified or identifiable from data as a key trigger for the application of data protection rules. This research explores the significance of the legal requirement that an individual must be identified or identifiable from data as tantamount to the primary factor in determining whether it is personal data or not and subject to EU data protection law, both now and under incoming legal reform (the EU General Data Protection Regulation, ‘GDPR’) to take effect from 25 May 2018. The research findings enable an assessment to be made of the ‘fitness-for-purpose’ of an identificatory approach to personal data as providing a meaningful boundary to data protection regulation. The yard-stick of evaluation is achievement of the twin aims of EU data protection legislation: upholding the fundamental rights of data subjects (in particular, their right to privacy) in connection with the processing of data about them to a high level of equivalence in EU national laws, while also facilitating intra-EU / transborder-with-EU flows of such data. Such an approach is compared against the contours of a new theoretical approach to personal data classification revolving around an analysis of the likely negative effects (risk of harm) to a data subject flowing from data processing activities intended to be applied to information. This comparison is then elaborated on via interrogation of the two approaches, and models founded upon such approaches (existing and prospective), in relation to determining when personal data may – and may no longer be - deemed personal under data protection law. This thesis argues that reconceptualising the definition of personal data as dependent on an effects-based assessment, not an identificatory one, could lay the foundations for a more conceptually-coherent, jurisdictional methodology for determining when and, ultimately, what data protection rules should apply in context. Such a methodology would require a risk-management exercise to be carried out by those planning to process data relating to persons, involving the quantification of likely harm that may flow from particular data processing activities in context. Consequently, a proposition is advanced to implement effect-based exemptions under data protection law, which takes insight from an effects-based framework already evolved under modernised EU competition law. In particular, consideration is given to the use of certain legal ‘safe harbour’ instruments - specifically, block exemption regulations providing comfort from enforcement action - as inspiration for moving to a more coherent, flexible and proportionate regulatory system able to take in to account collective interests. The resulting expansion of the data protection regulatory toolkit could incentivise and enhance confidence in compliance, as well as help encourage data sharing to promote innovation and demonstrable public benefits. Such a development would marry well with the incoming principles of accountability, and impact assessments, as the regulatory ‘lynchpins’ against which those who intend to process data relating to living individuals must abide by in the future as analytic practices and technological applications become more complex.

Towards A New Approach To The Legal Definition Of Personal Data And A Jurisdictional Model Of Data Protection Law: Surpassing The Requirement For An Assessment Of Identifiability From Data With An Effects-Based Approach

Piercing the Anonymity Veil: Re-identification Risk and the UK Transparency Agenda

The legal and privacy toolkit is a crucial component of the Data Pitch project (the Project). The aim for this first version of the toolkit is to address the question - “What are the critical things to consider when sharing and reusing data for defined innovative purposes under the Project?” The focus of the toolkit is on ensuring that legal rights in relation to shared data are respected, and legal obligations in respect of such data are followed in the Project. In particular, the toolkit provides information on how personal data will be anonymised before it is shared for analysis by SMEs taking part in the Project, and will describe techniques to mitigate the risk of harm befalling data subjects as a consequence of data relating to them being processed in new ways (for secondary purposes). As a project deliverable, this toolkit also provides an overview of the legal and regulatory framework that applies to data sharing and data reuse, including setting out key considerations governing the data sharing arrangements between the parties involved in the Project. The toolkit provides a ‘map’ of the relevant legal issues arising in the context of data sharing, and an outline of a methodology for handling these in a suitably risk-averse manner. This framework aims to treats data ethically and responsibly, with comprehensive, yet pragmatic guidance on data disclosure and its handling. The need for a toolkit under the Project is against a backdrop of emerging legal complexities in handling and repurposing data in a ‘big data’ environment in ways that are fair. This is especially important where personal data are to be subject to secondary processing. The vast power of data analytics means that absolute anonymisation of personal data may not be possible, so organisations have to focus on mitigating the risks associated with data sharing to the point where they are remote. Finding the correct balance between innovation and data protection without running the risks of over-reaction can be extremely difficult. This toolkit sets out a structured strategic approach to managing those potential issues that arise upfront, our goal is to make it easier to innovate safely and fairly. In this way, the toolkit hopes to present solutions to achieve legally-compliant data use in technically-enhanced and risk- aware ways, which also allows organisations to obtain high levels of utility from analysing shared datasets not currently in the public domain and hence to bring better services, products and research to the public.

Legal and Privacy Toolkit v1.0

Popular Tasks used by Researchers

How can I help with your research?

Get more enhanced response by uploading the PDFs you want me to reference.

No relevant PDFs in your library

SciSpace is the AI research assistant for academics. Run systematic literature reviews on 280M+ papers, and write papers with cited sources. Trusted by 1M+ students, PhDs & researchers.

SciSpace | AI for Research

We encountered an error while loading the panel. Please try refreshing the page.

Something went wrong

Analyze PDFs

Code & Manuscripts

Funding & Grants

Literature & Patents

Medical & Clinical Data

Systematic Review

Visualize & Present

Web & Data

Build a Google Scholar-like website for your research.

Build a website

Create charts and images for your research

Create a Chart

Write a paper for submission to a journal

Draft a manuscript

Patent Search

Design eye-catching scientific posters in minutes.

Scientific Poster Generation

Systematic Literature Review

Anyone with the link can view this chat. You can switch it back to private anytime.

One task is running at the moment. Your messages will be shown right after.

Drag and drop or click here to browse

Loved by <highlight>1 million+</highlight> researchers

Extract a list of specific topics and their sources from unstructured text

Topics

Compare and analyze relevant papers that matches with your search

Papers

Get insights from PDFs and bookmarked papers from your library

My library

Recent searches

Try searching for:

Catch AI-generated content in scholarly and non-scholarly content

{ai} Detector

Ai Writer

Get PDF Summaries, highlighted text explanations 

Chat with PDF

Effortlessly create in-text citations and bibliographies in APA and 2,500 other formats

Citation generator

Get explanations, summaries, and answers on academic papers

Ease up your research workflow with {scispace}'s cohort of exciting AI tools

Elevate your academic writing skills and convey your ideas the way you want

Paraphraser

Explore our range of reading and writing tools

Your file is being prepared and should be ready in a few minutes. If it's a large file, it might take a bit longer. You can close this window, and we'll email you the file when it's done.

You have reached a maximum limit of <strong>{limit}</strong> columns in the table. Remove at least <strong>1</strong> column to add or create another one.

Alison Knight

Author Tools

Chat about Author