Introduction:The study examined blockchain technology as a pillar of Web3, highlighting its principles of immutability, transparency, and decentralization. It analyzed the paradox that these same virtues could become disadvantages when it was necessary to correct errors, delete data, or deal with malicious uses, generating legal and ethical tensions.Development:Cases and studies were reviewed that showed how immutability guaranteed integrity and resistance to censorship but was incompatible with rights such as the “right to be forgotten” under the GDPR. Situations were also documented in which decentralization empowered both legitimate actors and criminals, eliminating consumer protection mechanisms. Faced with these dilemmas, solutions such as off-chain storage, updatable smart contracts, decentralized identity, and zero-knowledge proofs were evaluated. The proposal for double validation was highlighted, which incorporated a layer of smart contract verification to authenticate the origin and legitimacy of information before it was recorded. The validation of sensitive content by the people involved was also proposed as a strategy to prevent defamation, misinformation, or the dissemination of illegal material.Conclusion:The paper concluded that the potential of blockchain lay in its integration within an ethical, legal, and social framework. The implementation of mandatory verification and validation mechanisms strengthened accountability and individual protection, transforming blockchain into a tool that is not only secure and transparent, but also fair and socially responsible.
As generative AI (GenAI) technologies proliferate in urban governance, the challenge of building trustworthy AI systems becomes increasingly urgent. This chapter critically examines “trustworthiness” not as a purely technical attribute, but as a socio-political construct shaped by power, participation, and policy. Focusing on smart cities as testbeds of algorithmic governance, it explores how decentralized Web3 technologies—such as blockchain, DAOs, and data cooperatives—can offer structural alternatives to centralized, opaque systems. Drawing on action research from the Horizon Europe ENFIELD project and framed by EU policy developments like the AI Act and the Draghi Report, the chapter proposes a multi-layered governance model. It evaluates seven emerging techniques to strengthen GenAI accountability: (i) federated learning, (ii) blockchain provenance tracking, (iii) zero-knowledge proofs, (iv) DAO-based verification, (v) digital watermarking, (vi) explainable AI (XAI), and (vii) privacy-preserving machine learning (PPML). The chapter ultimately argues that trustworthy AI must be embedded in participatory governance, algorithmic transparency, and plural civic oversight. By reframing trust as a relational, institutional, and democratic issue, it contributes to reimagining smart cities not as technocratic projects, but as inclusive arenas for data justice and democratic renewal.
Contemporary Artificial Intelligence ("AI") systems, particularly Large Language Models ("LLMs"), face an imminent shortage of high-quality, humangenerated textual data, a phenomenon often termed "data exhaustion".This article examines the limitations of existing centralized data-annotation frameworks, highlighting critical issues such as bias, high computational overhead, and insufficiently adaptive infrastructures.Current market participants-including Scale AI, Appen, CloudFactory, and others-excel at rapidly scaling annotation services yet struggle with ethical sourcing, privacy compliance, and equitable compensation.In addition, legal and regulatory concerns, exemplified by stringent mandates such as the General Data Protection Regulation ("GDPR"), constrain the free flow of data essential for advanced AI research.As a corrective measure, decentralized data production paradigms are proposed, including the adoption of smart contracts, token-based incentives, and participatory governance through Decentralized Autonomous Organizations ("DAOs").While existing decentralized initiatives-SingularityNET, Fetch.ai,Ocean Protocol, Numeraire, and DcentAI-offer incremental innovations in reputation management and stakeholder engagement, they fail to fully address the nuanced requirements of large-scale "Mechanical Turk"-style data creation.In contrast, the author proposes a Weighted Directed Acyclic Graph ("WDAG") governance model which provides a multi-dimensional reputation framework, facilitating real-time validation of data contributions, adaptive ethical and legal compliance, and collaborative oversight by diverse community members.Findings suggest that such WDAGcentric systems can more effectively maintain data quality, ensure ethical alignment, and incentivize broad participation, thereby mitigating the looming data shortage and expanding AI's societal benefits.Ultimately, successful implementation requires coordinated efforts among policymakers, industry practitioners, and civil society actors to sustain both the technological and ethical integrity of AI research.By integrating WDAG-based governance with emerging decentralized solutions, the AI community may realize a more equitable, scalable, and future-ready paradigm for data provisioning.
Open access
2 source records
Ethics and Social Impacts of AI
Artificial Intelligence in Healthcare and Education
Stefano Balietti, Pietro Saggese, Stefan Kitzler, Bernhard Haslhofer
This chapter explores how Decentralized Autonomous Organizations (DAOs), a novel institutional form based on blockchain technology, challenge traditional centralized governance structures. DAOs govern projects ranging from finance to science and digital communities. They aim to redistribute decision- making power through programmable, transparent, and participatory mechanisms. This chapter outlines both the opportunities DAOs present, such as incentive alignment, rapid coordination, and censorship resistance, and the challenges they face, including token concentration, low participation, and the risk of de facto centralization. It further discusses the emerging intersection of DAOs and artificial intelligence, highlighting the potential for increased automation alongside the dangers of diminished human oversight and algorithmic opacity. Ultimately, we discuss under what circumstances DAOs can fulfill their democratic promise or risk replicating the very power asymmetries they seek to overcome.
This article explores the transformation of the state’s role in regulating personal data in the post-GDPR world. The author analyzes the impact of the EU’s General Data Protection Regulation (GDPR) on the evolution of the global privacy protection landscape, identifying trends towards harmonization and fragmentation of national legislations. The changing functions of the state as a regulator and guarantor of personal data protection in the context of digitalization are unveiled. The potential of blockchain technologies and distributed ledgers in ensuring user control over data is investigated. The influence of the development of the data market and new business models on the regulatory approaches of states and corporations is analyzed. The consequences of the spread of decentralized services for the relationships between the state, business, and civil society are considered. Priority directions for improving Ukrainian legislation in the field of personal data protection are substantiated, taking into account the realities of Web 3.0 and the need to balance innovation and security. The key idea is that the post-GDPR world stands at a crossroads between further fragmentation of the regulatory landscape and a long path towards harmonizing privacy standards. The choice of development trajectory depends on the coordinated political will of states, corporations, and global civil society to protect personal data as a shared value that unites humanity in the digital age. The article delves into the complex interplay of technological, legal, and societal factors shaping the future of data governance, offering insights into the challenges and opportunities ahead. It highlights the need for adaptive and inclusive regulatory frameworks that balance individual rights, economic interests, and public goods in an increasingly data-driven world.
Sarfaraz Gudumian, A. Lizy, S Jagadeeswari, S. Chinnadurai · 6 authors
Blockchain technology’s secure, decentralized platforms have revolutionized multiple industries. This paper discusses possible problems with employing blockchain technology to ensure security. This poses a question about scalability, privacy, and regulatory compliance. It suggests an approach to Blockchain-based Digital Signature Security Analysis (B-DSSA). This solution increases digital signature algorithms using the transparency and immutability of blockchain. This engineering makes it possible for electronic communications to be valid, intact and non-repudiated, making it applicable for secured settings like healthcare, finance, supply chain management etc. As results indicate, substantial advances have been made in preventing unauthorized access and tampering in digital transactions. By combining public and private blockchains, this method achieves scalability while protecting the privacy of sensitive data. This configuration makes it possible for real-time applications in healthcare and finance by optimizing resource utilization, maintaining high data integrity, and enabling speedier processing. Document signing systems, identity verification, and contract execution are some of the examples showing the flexibility and endurance of blockchain-based security solutions through B-DSSA. The paper reveals how blockchain technology may revolutionize the field of safety, leading to further research on marketing orientation issues.
Blockchain technology leverages a cryptographic system to provide secure and immutable storage of transaction histories within a decentralised framework. While various industries have demonstrated interest in integrating blockchain into their IT systems, concerns regarding accessibility, privacy, performance, and scalability persist. Permissioned blockchain frameworks offer a viable solution for securing confidential records. Extensive research has been conducted to explore the opportunities, challenges, application areas, and performance evaluations of different public and permissioned blockchain platforms. Given the sensitive nature of medical information, healthcare organisations must adhere to various legal obligations, including HIPAA regulations, to protect these data. Although navigating these requirements can be challenging, it is crucial for safeguarding the reputation of healthcare providers, maintaining patient trust, and avoiding legal repercussions. Permissioned blockchains represent decentralised digital ledgers tailored to collaborate among businesses and organisations. Their popularity has increased significantly in recent years, resulting in the availability of several leading options, such as Hyperledger Fabric, Corda, Quorum, and MultiChain. Each of these platforms presents its own set of advantages and disadvantages. Although blockchain technology remains relatively nascent in the permissioned realm, several factors warrant consideration when comparing these platforms. This study will review the existing landscape of blockchain technologies in healthcare applications and identify the research scopes. This research aims to determine how permissioned blockchain technology can effectively fulfil the requirements for managing healthcare data.
Autonomous AI agents present transformative opportunities and significant governance challenges. Existing frameworks, such as the EU AI Act and the NIST AI Risk Management Framework, fall short of addressing the complexities of these agents, which are capable of independent decision-making, learning, and adaptation. To bridge these gaps, we propose the ETHOS (Ethical Technology and Holistic Oversight System) framework, a decentralized governance (DeGov) model leveraging Web3 technologies, including blockchain, smart contracts, and decentralized autonomous organizations (DAOs). ETHOS establishes a global registry for AI agents, enabling dynamic risk classification, proportional oversight, and automated compliance monitoring through tools like soulbound tokens and zero-knowledge proofs. Furthermore, the framework incorporates decentralized justice systems for transparent dispute resolution and introduces AI specific legal entities to manage limited liability, supported by mandatory insurance to ensure financial accountability and incentivize ethical design. By integrating philosophical principles of rationality, ethical grounding, and goal alignment, ETHOS aims to create a robust research agenda for promoting trust, transparency, and participatory governance. This innovative framework offers a scalable and inclusive strategy for regulating AI agents, balancing innovation with ethical responsibility to meet the demands of an AI-driven future.
Mir Mehedi Rahman, Bishwo Prakash Pokharel, Sayed Abu Sayeed, Sujan Bhowmik · 6 authors
In the evolving landscape of cybersecurity, traditional information technology (IT) infrastructures often struggle to meet the demands of modern risk management frameworks, which require enhanced security, scalability, and analytical capabilities. This paper proposes a novel artificial intelligence (AI)–driven IT infrastructure backed by blockchain technology, specifically designed to optimize risk management processes in diverse organizational environments. By leveraging artificial intelligence for predictive analytics, anomaly detection, and data-driven decision-making, combined with blockchain’s secure and immutable ledger for data integrity and transparency, the proposed infrastructure offers a robust solution to existing challenges in risk management. The infrastructure is adaptable and scalable to support a variety of risk management methodologies, providing a more secure, efficient, and intelligent system. The findings highlight significant improvements in the accuracy, speed, and reliability of risk management, underscoring the infrastructure’s capability to proactively address emerging cyber threats. To ensure the proposed model effectively addresses the most critical issues, the Decision-Making Trial and Evaluation Laboratory (DEMATEL) technique will be used to analyze and evaluate the interrelationships among the existing critical factors. This approach evaluates the interrelationships and impacts of these factors, verifying the model’s comprehensiveness in managing organizational risk. This study lays the foundation for future research aimed at refining AI-driven infrastructures and exploring their broader applications in enhancing organizational cybersecurity.
Md. Hasibul Alam Ratul, Sepideh Mollajafari, Martín Wynn
Digital evidence plays a crucial role in cybercrime investigations by linking individuals to criminal activities. Data collection, preservation, and analysis can benefit from emerging technologies like blockchain to provide a secure, distributed ledger for managing digital evidence. This study proposes a blockchain-based solution for managing digital evidence in cybercrime cases in the judicial domain. The proposed solution provides the basis for the development of a new model that leverages a consortium blockchain, allowing secure collaboration among judicial stakeholders, while ensuring data integrity and admissibility in court. An extensive literature review demonstrates blockchain’s potential to create a more secure, efficient evidence management system. The proposed model was implemented in a test environment using a localised blockchain for developing and testing smart contracts, as well as integrating a web interface, with off-chain storage for managing evidence data. The system was subsequently deployed in both the Polygon and Ethereum test networks, simulating real-world blockchain environments, revealing that the operational cost in the Polygon network is reduced by 99.96% compared to Ethereum, thereby offering scalability without compromising security. This study underscores blockchain’s potential to revolutionise the chain of custody procedures, improving dependability and security in evidence management and providing more sustainable solutions within the criminal justice system.
Metaverses may present innovative channels for business and finance , education, and workplaces. Drawing from the Diffusion of Innovation Theory and the Unified Theory of Acceptance and Use of Technology (UTAUT) as well as the literature on digital equality, this research attempts to unravel the dynamics of digital equality and trust in AI-empowered metaverses for various industry sectors. Three cross-sectional surveys ( N Total = 1086) examined US Internet users' intention to adopt metaverses for business, education, and workplaces. Structural equation models were estimated using Mplus 8.8. Study 1 indicates dynamic relationships among digital equality (commerce dimension), blockchain transparency, privacy concerns, and adoption intention. Study 2 shows dynamic associations among digital equality (educational, social, and political dimensions), digital adaptability, loneliness, and adoption intention. Study 3 demonstrates dynamic interconnections among digital equality (labor, government, and health dimensions), digital adaptability, identity threat, and adoption intention. Across three datasets, trust mediates the relationship between digital equality and adoption intention. Theoretical contributions to the emerging literature on decentralized finance (DeFi), AI-driven digital transformation, and AI-VR-convergence are discussed. DeFi businesses, enterprises, policy makers, educators, professional training providers, and workforce developers need to consider third-level digital (in)equality and digital adaptability in developing equitable, sustainable, and inclusive user experience (UX) in AI-empowered metaverses.
This study aims to identify and assess AI and blockchain solutions in relation to journalistic authenticity and integrity. Central to our exploration is the role of blockchain technology in verifying content provenance. As a key component of a global Web3 framework, blockchain could offer a foundation for authenticating the origins of content. In this article, we explore how blockchain, with its capacity for creating immutable and cryptographically signed data records, could be applied by journalists to verify photos, videos and documents. Our analysis identified nine blockchain-based solutions for content verification, with three platforms–Attestiv, OriginStamp, and Fact Protocol–showing particular promise for journalistic workflows. We conclude that while AI and blockchain solutions are currently available to journalists today, they require high-level technical expertise. Many media companies are now venturing into this field as well, thus affecting the professional role of journalists in general. In our study, it is evident that integrating AI and blockchain in journalism is not merely about adopting new tools but also about understanding their broader implications for journalism as a profession and the convergence in society. The focus must remain on enhancing journalistic integrity and public trust to ensure that these technological advances benefit the field of journalism and, by extension, the democratic processes it supports.
Md. Rafiqul Islam, Muhammad Haseeb, Hina Batool, Nasir Ahtasham · 5 authors
The integrity of global elections is increasingly under threat from artificial intelligence (AI) technologies. As AI continues to permeate various aspects of society, its influence on political processes and elections has become a critical area of concern. This is because AI language models are far from neutral or objective; they inherit biases from their training data and the individuals who design and utilize them, which can sway voter decisions and affect global elections and democracy. In this research paper, we explore how AI can directly impact election outcomes through various techniques. These include the use of generative AI for disseminating false political information, favoring certain parties over others, and creating fake narratives, content, images, videos, and voice clones to undermine opposition. We highlight how AI threats can influence voter behavior and election outcomes, focusing on critical areas, including political polarization, deepfakes, disinformation, propaganda, and biased campaigns. In response to these challenges, we propose a Blockchain-based Deepfake Authenticity Verification Framework (B-DAVF) designed to detect and authenticate deepfake content in real time. It leverages the transparency of blockchain technology to reinforce electoral integrity. Finally, we also propose comprehensive countermeasures, including enhanced legislation, technological solutions, and public education initiatives, to mitigate the risks associated with AI in electoral contexts, proactively safeguard democracy, and promote fair elections.
Smart contracts form the core of Web3 applications. Contracts mediate the transfer of cryptocurrency, making them irresistible targets for hackers. We introduce ASP, a system aimed at easing the construction of provably secure contracts. The Asp system consists of three closely-linked components: a programming language, a defensive compiler, and a proof checker. The language semantics guarantee that Asp contracts are free of commonly exploited vulnerabilities such as arithmetic overflow and reentrancy. The defensive compiler enforces the semantics and translates Asp to Solidity, the most popular contract language. Deductive proofs establish functional correctness and freedom from critical vulnerabilities such as unauthorized access.
In museums, heritage, and non-profit cultural organisations, thought leadership on the ethical implications of AI is gathering speed. Notable initiatives include the Network of European Museum Organisations (NEMO)'s efforts to address the uptake of AI in museums (2024) and the UK's Arts and Humanities Research Council's BRAID programme dedicated to integrating Arts and Humanities research into the 'Responsible AI' ecosystem (2022). This is a fast-evolving area with new analysis and calls to action appearing with frequency. As yet however, little attention has been given to how AI is emotionally impacting lived experiences of cultural workers as organisations seek to operationalise it. This paper highlights the need to consider such impacts, as well as the general-purpose technologies AI builds upon, such as biotechnology and a connected ecosystem of devices, on cultural workers and their practices. The convergence of these technologies signals what futurist Amy Webb calls a technology "supercycle" with far-reaching implications (Aiello, 2024).This paper does not present new empirical evidence but rather portends the need for more research on the relationship between wellbeing, AI, and work in cultural organisations. Based in the UK and the Netherlands, the authors are researcher-practitioners who explore the impact of digital technology on cultural workforces. This paper critically anticipates the implications of AI on those working on the ground in cultural organisations. Our ideas and examples draw from the UK, Europe and the United States, but we hope they speak to experiences of cultural workers across the globe.In what follows, a strategic foresight method known as the "Futures Triangle" (Inayatullah, 2023) will be deployed to consider the plausible future of AI-driven cultural work that might emerge between three pushing and pulling cornersthe past, the present, and the futureeach of which is shaping the adoption of AI in our cultural organisations. This framework helps raise awareness of the trends, drivers, and signals of which cultural organisations need to be aware to ensure an integration of AI that empowers workers to engage effectively while addressing the ethical dilemmas involved. This structure recognises the history that has shaped our current decisions, determines what to carry forward or leave behind, and confronts present challenges. We must consider the impact of our choices and how they will resonate into the future, ensuring they reflect the people and context of multiple possible futures.In Future Thinking, trends refer to long-term patterns that can be carried forward into the future. There are two trends that have impacted cultural work, wellbeing, and AI: digital transformation and decolonisation.The recent history of digital transformation in museums, heritage and cultural organisations is convoluted. Many associate momentum to a 2017 review of English museums which argued for "dynamic data for dynamic collections" (DCMS, 2017: 64). The following year, the UK's Department for Culture, Media, and Sport (DCMS) launched Culture is Digital, a report calling for "practitioners and organisations across the cultural and tech sectors" (2018: 17) to develop "digital thinking" (ibid: 9) stating that organisations felt "held back" by a lack of infrastructure, resources, digital skills and leadership training, resulting in "a fragmented approach" to technology (ibid: 5). Despite digital developments in many cultural organisations the story remained patchy, with funding tending to prioritise "shiny" short-term over long-term infrastructure projects, an ongoing deficit of digital literacy, cataloguing backlogs, and a predominance of non-interoperable systems and formats. This often led to reluctance, fear, and apathy when it came to instilling digital change up to and since the Covid-19 pandemic. As a respondent in a survey on AI in heritage organisations stated: even if AI has the potential "to revolutionise the heritage sector", there remain "real problems that are affecting us, like skills shortages, funding shortages and volunteer shortages" (Oates, 2023). They continued: "the obsession with 'digital' is unhelpful, especially when digital is so poorly defined that the adjective is arbitrarily used as a meaningless noun" (ibid). However, as John Stack, Director of Digital Innovation and Technology at The National Gallery, notes: "Unlike previous recent waves of technology (crypto currency, web3, NFTs, etc.) the AI revolution feels different -there's a sense in which it seems likely to change many things, but we don't fully understand the potential and implications" (Stack, J. (2024) Email to Sophie Frost, 9 January).Trend 2: Decolonisation Often founded on colonial principles, cultural organisations carry legacies that have both hindered and guided their embrace of new technologies. Decolonisation initiatives have burgeoned in recent years with bodies such as UNESCO, International Council of Museums (ICOM), American Alliance of Museums, and Museums Association supporting the need to "recognise the integral role of empire in museumsfrom their creation to the present day" (Museums Association, 2024). While actions for enhanced traceability and diligence in the acquisition of cultural objects has grown, there has been less vigilance regarding the integration of emerging technologies which invite new, equally problematic, forms of coded gaze into legacy institutions. Despite conversations amongst museum and heritage practitioners on the dangers of commercial technologies (Pratty, 2019), there have been few practical attempts to source long-term ethical alternatives. As Oonagh Murphy, editor of this Issue, points out: there is a need to "engage in a broader conversation about the power and impact" of digital tools and products (2024: 73). NEMO has implied that museums have the "potential" to be "partners in the development of ethical practices related to emerging technologies" ( 2024), but there is more work to be done to link decolonisation efforts with the integration of AI.Drivers are broad long-term forces that are likely to have a significant impact on the future. Drawing on research exploring AI in the global workplace, we locate the following as shaping how we imagine the signals to emerge in the future.Research suggests that AI will predominantly result in augmentation rather than automation (International Labour Organization, 2024: 6). Clerical work will be most exposed to automation because of generative AI, with 24% of tasks "highly exposed" (ibid). This is a gendered issue. As women tend to be overrepresented in the clerical field, women's jobs may be "twice as likely to be exposed" as men's (Muldoon et al., 2024: 53). Museum and heritage environments have long been recognised as "pink collar" workplaces (GEMM, 2019) whereby women undertake most administrative roles. In the USA, 58.6% of the workforce is female across archival, curatorial and museum technician roles (DataUSA, 2022) while a UK report on the arts, culture and heritage workforce found only 34% of women occupy managerial or director positions with the majority in junior roles (PEC, 2024). Jobs in areas such as collections management could have large components of their roles augmented by AI. Large Language Models excel at "formal, standardised tasks with clear objectives and large amounts of text data" (Muldoon et al. 2024: 53). In cultural organisations, this form of digital data entry is typically done by women on relatively low pay (Frost, 2022). Generative AI is also valuable in content creation, meaning that marketing and social media roles may be affected.Driver 2: Demand for AI Literacy Discussions of AI workplace integration emphasise reskilling workforces to support its effective use. In the context of cultural organisations, stakeholders are recognising AI literacy as an opportunity to reshape digital work, introduce new efficiencies, whilst emphasise the value of employee agency to digitally experiment through AI methods. Industry specialist Jocelyn Burnham, who offers AI workshops for the cultural and creative sectors, recognises AI as a tool for experimenting with what new technologies might offer, rather than being anxious about how they might disrupt the status quo. AI prompts us to ask more critically engaged questions, to understand our own needs and the needs of audiences in different ways. Large Language Models (LLMs) are helping workforces augment their digital skills, enabling those who have not been able to use new technologies so easily in the past to do so, whilst helping staff learn in different ways (Burnham, J. (2024) Interview with Sophie Frost, 17 January 2024).There is a flipside. The International Labour Organization notes that some parts of the world are at risk of an "AI divide" whereby "high income nations disproportionately benefit from AI advancements" (2024: 5). This could be the case in the cultural field also where AI training is inconsistent. As Angie Judge, CEO of Dexibit, notes: "AI will quickly create a world of the haves and have nots. I hope the museum sector will find itself on the right side of that equation" (Styx, 2024). For Stack, "best practice is still emerging and much of the work is 'bottom up' with museum practitioners exploring the potential, rather than top down with managers and leaders directing this work". He continues: "museums are starting to recognise the need for a policy document that is set for internal review every six months or so, while other internal museum policies often go years before review" (Stack, J. ( 2024) Email to Sophie Frost, 9 January). Such need for agility in the creation of AI policy has been recognised by the UK government also: their Generative AI Framework for HMG describes itself as "necessarily incomplete and dynamic" (Gov.uk, 2024).The most consequential driver for cultural organisations may be the inherent bias embedded within and potential misuse of LLM training data. Many have documented the ways AI datasets have been drawn through the rules and algorithms of those who trained them, leading to race and gender discrimination across multiple online platforms (Leavy et. al., 2020;Buolamwini, 2023). Much of this software has been invested in by legacy tech companies and it is widely recognised that technological development is under the control of those "with strong imperatives to continue expanding their operations and increasing their profits" (Muldoon et al. 2024: 161). New government guidelines for the public sector emphasise the need, when working with AI, "to establish and communicate how you will address ethical concerns from the start" (Gov.uk, 2024) but how can smaller, less powerful organisations reclaim power when limited by the options available to them? Helpfully, the European Commission's recent foresight report focused on the future of Big Tech in Europe and its implications for research and innovation (R&I). Using a scenario-based approach, it offers recommendations to guide the EU's R&I policy. Significantly, all four imagined scenarios indicated "varieties of high tech capitalism" with only one emphasising the prosperity of civil society over Big Tech (2024: 8). Cultural organisations can use such resources to understand the current landscape and strategically respond not only to present developments but anticipate alternative futures of information control.There is an additionally troubling aspect regarding the increased risk of copyright infringement for cultural organisations who make income from their picture libraries through licensing for reproduction and commercial research. This issue gained attention in the UK through high-profile publications such as the House of Commons Culture, Media and Sport Committee report on Content Remuneration, which calls on government "to ensure that creators have proper mechanisms to enforce their content and receive fair compensation when their works are used by AI systems" (House of Commons, 2024). If this driver is not managed proactively, what might be its effects on income in years to come?The exponential growth of AI-related employment has surfaced questions regarding its hidden human costs (Muldoon, Graham & Cant, 2024;Gray & Suri, 2022;Catanzariti et al., 2021). Attention is being paid to how human labour plays a pivotal role in enabling AI technology, particularly generative AI, through the paid, piecework of collecting, processing, and labelling of datasets needed by models such as ChatGPT for training. Our research has observed the emotional toll of digitally driven labour in museums, heritage, and cultural organisations (Frost, 2021 and2022;Vargas, 2020). Daily tasks such as establishing suitable naming conventions, cleaning up and figuring out problems in the data, and dealing with acquisition backlogs, require persistence, care, clarity, and impartiality. Those who promote digital change experience personal, psychological costs and ethical dilemmas consistent with other types of cultural and creative work (see Banks, 2017;Belfiore, 2021). When it comes to AI, the concern of many digital staff is, as Steven Franklin, Social Media Manager at The Royal Institution explains, "if you've got technology that is inherently designed to increase individuals' productivity then the demands of the job will probably go up with it. So, you're going to be expecting more people to do more" (Franklin, S. ( 2024) Interview with Sophie Frost, 12 January).A recent study explored the "dark side effects of digital working" of 142 workers, specifically their levels of stress, overload, anxiety and Fear of Missing Out on Information (Marsh et. al., 2024). It found that "employees who are overloaded by information or worried about missing out on it in the digital workplace face risks to their well-being at work" (2024: 12) and stated that greater consideration of the digital workplace "is essential to not only employee productivity but also wellbeing in modern organisations" (ibid). Cultural workplaces similarly need to reflect on how AI will impact well-being, and how the information overload of working with new technology has emotional consequences for workers.It is easy to suggest that AI poses a threat to those who have spent years honing their skills as cultural workers, drawing on data suggesting that those with higher qualifications are more exposed to AI (Department for Education 2023: 18). We propose instead that cultural roles will be positively augmented by AI, creating opportunities for better discoverability and searchability in daily operations as well as in creative reuse and reimagination within the interpretation of collections and audience engagement. AI will need to be accommodated and understood within all job roles in the cultural workplace; it will become the responsibilityand the possibility -of all (and yes, this will require consistent resourcing of time, people, energy, and moneyas does all digital maturity work).More concerning for us is a wider existential issue: if AI, as has been argued by many in the creative industries, is devaluing creativity as a specifically human endeavour, will this deprioritise the importance of the institutions that care for it? We have heard outcries, in areas such as screenwriting, music production, video games and the visual arts regarding the use of generative AI and other digital technologies to replicate human creative outputs (SAG-AFTRA, 2024;Weiss, 2023;The Art Newspaper, 2023). As the quality of generative AI develops, will there not be a moment when even experts are unable to distinguish between AI and human generated cultural artefacts? What might this mean for history itself, for the historical record? Nick Bostrom's Deep Utopia ( 2024) offers one answerhe invites us to rethink the complexity of utopia and confront the unintended consequences of our pursuit for better representation and justice, challenging the binary thinking of utopian vs. dystopian outcomes. He argues that true progress instead requires moving beyond fixed notions of perfection to consider multifaceted possibilities and risks of future realities.Signals represent small, local innovations or disruptions that have the potential to grow in scale and distribution (Howard, 2021). Below are five signals, posed as 'What if?' questions, that critically anticipate how cultural workplaces might tackle the present and reimagine a future with AI. "What if" questions encourage creative thinking, challenge assumptions, and open alternative possibilities. For futurist Peter Schwartz, they are part of scenario planning, a strategic tool to prepare for uncertainty by imagining diverse future scenarios (1991).1. What if we developed greater "coopetition" -cooperating with our competitors to achieve a common goal -as we integrate AI in our cultural workplaces? Our sector must radically cooperate with other stakeholders on the ethics of AI -from corporations to governments, from other cultural organisations to grassroots bodies.workplaces? The integration of AI in our cultural workplaces has required on-the-job learning. New technological developments mean greater demand for new digital literacies in our workplaces; it is imperative to embrace lifelong learning and the upskilling of workforces.technology, climate change, and social justice, and acknowledged more openly how they in our cultural workplaces? are trends and thinking about can reshape and of and What if we that the emotional and psychological of digital work to employment practices in our cultural workplaces? the wellbeing of cultural workers and and by recognising the human AI, cultural organisations will ensure that AI both work and public What if we the enabling AI, their impact on cultural workers, our role as cultural to the global conversation on AI in ways that reflect and lived experience in our cultural workplaces? This encourage us to the AI development and consider how cultural can their to guide practices that the future of the cultural is needed is more empirical research on how employee wellbeing and job quality in to AI innovation are being in museums, heritage and cultural more with other parts of the cultural and creative as we operationalise and more at the of leadership and -of how AI and within a of and cultural we need to learn to better and this ensuring that the integration of AI and lived experiences of cultural workers rather than
Caspar Barnes, Mateo Aboy, Timo Minssen, Jemima Winifred Allen · 7 authors
Participation in research is supposed to be voluntary and informed. Yet it is difficult to ensure people are adequately informed about the potential uses of their biological materials when they donate samples for future research. We propose a novel consent framework which we call "demonstrated consent" that leverages blockchain technology and generative AI to address this problem. In a demonstrated consent model, each donated sample is associated with a unique non-fungible token (NFT) on a blockchain, which records in its metadata information about the planned and past uses of the sample in research, and is updated with each use of the sample. This information is accessible to a large language model (LLM) customized to present this information in an understandable and interactive manner. Thus, our model uses blockchain and generative AI technologies to track, make available, and explain information regarding planned and past uses of donated samples.
Open access
Ethics in Clinical Research
Artificial Intelligence in Healthcare and Education
This paper presents an approach for the verification of access control in smart contracts written in the Digital Asset Modeling Language (DAML). The approach utilizes Colored Petri Nets (CPNs) and their analysis tool CPN Tools. It is a model-driven-based approach that employs a new meta-model for capturing access control requirements in DAML contracts. The approach is supported by a suite of tools that fully automates all of the steps: parsing DAML code, generating DAML model instances, transforming the DAML models into CPN models, and model checking the generated CPN models. The approach is tested using several DAML scripts involving access control extracted from different domains of blockchain applications.
The vision of Web3 is to improve user control over data and assets, but one challenge that complicates this vision is the prevalence of non-transparent, scam-prone applications and vulnerable smart contracts that put Web3 users at risk.While code audits are one solution to this problem, the lack of smart contracts source code on many blockchain platforms, such as Sui, hinders the ease of auditing.A promising approach to this issue is the use of a decompiler to reverse-engineer smart contract bytecode.However, existing decompilers for Sui produce code that is difficult to understand and cannot be directly recompiled.To address this, we developed the SuiGPT Move AI Decompiler (MAD), a Large Language Model (LLM)-powered web application that decompiles smart contract bytecodes on Sui into logically correct, human-readable, and recompilable source code with prompt engineering.Our evaluation shows that MAD's output successfully passes original unit tests and achieves a 73.33% recompilation success rate on real-world smart contracts.Additionally, newer models tend to deliver improved performance, suggesting that MAD's approach will become increasingly effective as LLMs continue to advance.In a user study involving 12 developers, we found that MAD significantly reduced the auditing workload compared to using traditional decompilers.Participants found MAD's outputs comparable to the original source code, improving accessibility for understanding and auditing non-open-source smart contracts.Through qualitative interviews with these developers and Web3 projects, we further discussed the strengths and concerns of MAD.MAD has practical implications for blockchain smart contract transparency, auditing, and education.It empowers users to easily and independently review and audit non-open-source smart contracts, fostering accountability and decentralization.Moreover, MAD's methodology could potentially extend to other smart contract languages, like Solidity, further enhancing Web3 transparency.
Many smart contracts are prone to exploits, which has given rise to analysis tools that try to detect and fix vulnerabilities. Such analysis tools are often trained and evaluated on limited data sets, which has the following drawbacks: 1. The ground truth is often based on the verdict of related tools rather than an actual verification result; 2. Data sets focus on low-level vulnerabilities like reentrancy and overflow; 3. Data sets lack concrete exploit examples. To address these shortcomings, we introduce XploGen, which uses a model-based oracle specification of the business logic of the smart contracts to synthesize valid exploits using LLMs. Our experiments, involving 104 synthesized vulnerability-exploit pairs, demonstrated a 57% success rate in exploiting targeted aspects of the contract. They achieved exploit efficiency with an average of only 3.5 transactions per exploit, highlighting the effectiveness of our methodology.
Since funds or tokens in smart contracts are maintained through specific state variables, contract audit, an effective means for security assurance, particularly focuses on these variables and their related operations. However, the absence of publicly accessible source code for numerous contracts, with only bytecode exposed, hinders audit efforts. Recovering variables and their types from Solidity bytecode is thus a critical task in smart contract analysis and audit, yet this is a challenging task because the bytecode loses variable and type information, only with low-level data operated by stack manipulations and untyped memory/storage accesses. The state-of-the-art smart contract decompilers miss identifying many variables and incorrectly infer the types for many identified variables. To this end, we propose VarLifter , a lifter dedicated to the precise and efficient recovery of typed variables. VarLifter interprets every read or written field of a data region as at least one potential variable, and after discarding falsely identified variables, it progressively refines the variable types based on the variable behaviors in the form of operation sequences. We evaluate VarLifter on 34,832 real-world Solidity smart contracts. VarLifter attains a precision of 97.48% and a recall of 91.84% for typed variable recovery. Moreover, VarLifter finishes analyzing 77% of smart contracts in around 10 seconds per contract. If VarLifter is used to replace the variable recovery modules of the two state-of-the-art Solidity bytecode decompilers, 52.4%, and 74.6% more typed variables will be correctly recovered, respectively. The applications of VarLifter to contract decompilation, contract audit, and contract bytecode fuzzing illustrate that the recovered variable information improves many contract analysis tasks.
Shelly Grossman, John Toman, Alexander Bakst, Sameer Arora · 6 authors
SMT-based verification of low-level code requires modeling and reasoning about memory operations. Prior work has shown that optimizing memory representations is beneficial for scaling verification—pointer analysis, for example can be used to split memory into disjoint regions leading to faster SMT solving. However, these techniques are mostly designed for C and C++ programs with explicit operations for memory allocation which are not present in all languages. For instance, on the Ethereum virtual machine, memory is simply a monolithic array of bytes which can be freely accessed by Ethereum bytecode, and there is no allocation primitive. In this paper, we present a memory splitting transformation guided by a conservative memory analysis for Ethereum bytecode generated by the Solidity compiler. The analysis consists of two phases: recovering memory allocation and memory regions, followed by a pointer analysis. The goal of the analysis is to enable memory splitting which in turn speeds up verification. We have implemented both the analysis and the memory splitting transformation as part of a verification tool, CertoraProver, and show that the transformation speeds up SMT solving by up to 120× and additionally mitigates 16 timeouts when used on 229 real-world smart contract verification tasks.
Electronic health records (EHRs) are increasingly replacing traditional paper-based medical records due to their speed, security, and ability to eliminate redundant data. However, challenges such as EHR interoperability and privacy concerns remain unresolved. Blockchain, a distributed ledger technology comprising connected, encrypted data blocks, presents a promising solution. This study explores how blockchain technology can revolutionize hospital EHR management. Our proposed solution securely transfers medical records between patients and doctors using the InterPlanetary File System (IPFS) and the Ethereum platform. Utilizing smart contracts automates data transfers, ensuring patient anonymity and reducing computational complexity while securely storing patient data on the network. Patient records are stored locally on the Ganache server, with the front end managed using HTML, CSS, ReactJS, and JavaScript, and the backend developed in Solidity. Blockchain technologies combined with Role- Based access control instead of attribute -based access control. The system's throughput increases linearly with the number of users and requests, enhancing the framework's efficiency and scalability. The minimum recorded latency is 14 ms.
How AI models should deal with political topics has been discussed, but it remains challenging and requires better governance. This paper examines the governance of large language models through individual and collective deliberation, focusing on politically sensitive videos. We conducted a two-step study: interviews with 10 journalists established a baseline understanding of expert video interpretation; 114 individuals through deliberation using InclusiveAI, a platform that facilitates democratic decision-making through decentralized autonomous organization (DAO) mechanisms. Our findings reveal distinct differences in interpretative priorities: while experts emphasized emotion and narrative, the general public prioritized factual clarity, objectivity, and emotional neutrality. Furthermore, we examined how different governance mechanisms - quadratic vs. weighted voting and equal vs. 20/80 voting power - shape users' decision-making regarding AI behavior. Results indicate that voting methods significantly influence outcomes, with quadratic voting reinforcing perceptions of liberal democracy and political equality. Our study underscores the necessity of selecting appropriate governance mechanisms to better capture user perspectives and suggests decentralized AI governance as a potential way to facilitate broader public engagement in AI development, ensuring that varied perspectives meaningfully inform design decisions.