Voluntary Codes Research Articles

Large language models (LLMs): Risks and policy implications

"Although statistical language modeling has been a fairly active area of academic research in artificial intelligence (AI) since the 1980s, the recent explosion of both technological innovation and public attention around LLMs has dramatically changed the space of applications for this work. The unprecedented scale of these new models carries with it emergent capabilities—particularly in the form of generative models—that have made the technology generalizable to a broad range of use cases and accessible to the general public. Yet these rapidly developing features also pose novel risks, ranging from environmental impacts and copyright violations to misinformation and hate speech. Traditional regulatory frameworks are not well-equipped to match either the rapid pace of innovation or the ability of LLMs to transcend the jurisdictions of individual governing bodies. Thus far, the governance gap has been filled by soft law frameworks such as private company standards, voluntary codes of conduct, and design guides from non-regulatory standard-setting organizations. Still, there is rising demand to solidify these soft guidelines into hard law. In addition to offering a historical overview of LLMs and highlighting some of the most pressing concerns involving LLMs today, we discuss legislative attempts to address these concerns and outline potential complicating factors."

Sustainable Consumption and Social Justice: Should Countries Take Action to Ban the Sale of Products Made under Sweatshop Conditions?

This paper investigates the subject of sweatshop labour and the debate over whether countries should prohibit the sale of products produced under such conditions. The introduction emphasises the importance of better working conditions and minimal standards in sweatshop factories, while also admitting that customers may be ready to pay slightly higher costs to support better working circumstances for sweatshop workers. However, economists say that sweatshop labour is preferable than not working or working for extremely low wages, and that a blanket ban on sweatshop products could result in job losses and other negative repercussions. The literature study explores a range of viewpoints on sweatshop labour, including moral and financial grounds. Additionally, it highlights how consumer movements and market pressure influence businesses to improve working conditions. The paper also emphasises the issue's complexity and the need for a fair policy that considers employees' rights and practical factors. To improve the working circumstances of sweatshop workers, a multi-pronged strategy is recommended. This method includes consumer pressure, government regulation, and voluntary codes of conduct. The importance of striking a compromise that takes into account differing points of view and deals with the complexities of the sweatshop labour issue is emphasised in the conclusion. Overall, the study provides a thorough assessment of the subject, emphasising the necessity for ethical manufacturing processes and improved working conditions while taking into account economic ramifications and unexpected consequences.

The anti-politics of impact investment: Financial self-regulation, market competition and over-indebtedness in Cambodia

Social impact investors claim to promote sustainable development by mobilizing private finance capital to solve pressing global challenges like poverty alleviation. In this paper, I interrogate this claim through an examination of microfinance in Cambodia, a major destination for impact investment. In the past decade, Cambodia has received nearly 10% of global investments from microfinance-specific funds. It now has the largest microfinance-debt per capita ratio of any country in the world. Based on qualitative research between 2021 and 2023, I argue that impact investment functions as an anti-politics of development that reinforces finance as a political technology of neoliberal governance. Over the past two decades, impact investors have poured capital into the country’s microfinance industry to expand access to credit without acknowledging structural political-economic conditions that have produced rising over-indebtedness among microfinance borrowers. Instead, they have argued that the problems created by the microfinance industry are best resolved through a self-regulation model that uses a voluntary code of conduct based on global standards of responsible finance. Thus, impact investors have been integral in broader transformations that have extended financial logics, technologies and accumulation imperatives into people’s daily lives. This paper contributes to economic geography and critical development studies by explaining how impact investment deepens neoliberal financialization.

Global Regulatory Frameworks for the Use of Artificial Intelligence (AI) in the Healthcare Services Sector.

The healthcare sector is faced with challenges due to a shrinking healthcare workforce and a rise in chronic diseases that are worsening with demographic and epidemiological shifts. Digital health interventions that include artificial intelligence (AI) are being identified as some of the potential solutions to these challenges. The ultimate aim of these AI systems is to improve the patient's health outcomes and satisfaction, the overall population's health, and the well-being of healthcare professionals. The applications of AI in healthcare services are vast and are expected to assist, automate, and augment several healthcare services. Like any other emerging innovation, AI in healthcare also comes with its own risks and requires regulatory controls. A review of the literature was undertaken to study the existing regulatory landscape for AI in the healthcare services sector in developed nations. In the global regulatory landscape, most of the regulations for AI revolve around Software as a Medical Device (SaMD) and are regulated under digital health products. However, it is necessary to note that the current regulations may not suffice as AI-based technologies are capable of working autonomously, adapting their algorithms, and improving their performance over time based on the new real-world data that they have encountered. Hence, a global regulatory convergence for AI in healthcare, similar to the voluntary AI code of conduct that is being developed by the US-EU Trade and Technology Council, would be beneficial to all nations, be it developing or developed.

Co‐production of agroecological innovations to improve sustainability in South American fruit farms

Abstract Agricultural intensification and expansion are the main drivers of biodiversity loss that continue to increase this century, especially in South America. International markets and global policy provide incentives and frameworks to address this, but these are unlikely to be effective unless farmers on the ground are enabled and motivated to respond to them by developing long‐term solutions that fit their production systems and local contexts. Here, we use a multi‐actor transdisciplinary approach to co‐design and test agroecological innovations suitable for intensive, exporting South American fruit farms. We focus on highly biodiverse regions experiencing habitat loss in the Mediterranean and dry tropical forest regions of Chile and Brazil, respectively. The innovations were designed to support local biodiversity without compromising productivity or quality. Fourteen farmers participated throughout the project, covering a total of 4178 ha of intensive table grape, mango and cherry production. All were under pressure from buyers to report action on biodiversity. Farmers worked with researchers and industry representatives through an iterative process of dialogues and workshops to select, co‐design and implement three agroecological innovations: perches for birds of prey, cover crops and native hedgerows. Farmers became engaged in monitoring their effectiveness and redesigning them to suit local contexts. We develop an extensive set of resources for ongoing dissemination, including an online sustainability metric to report the practices carried out. Eight farms continued to implement at least one agroecological innovation beyond the end of the project, motivated by its fit to their management system and their ability to report positive actions in their supply chains. Policy implications. Our model of knowledge co‐production demonstrates how transdisciplinary research in agriculture, fully localised in a particular food‐producing context, can enable farmers in the global South to engage with biodiversity conservation in response to top‐down market signals incentivising sustainability. We argue that many top‐down efforts to enhance the sustainability of food supply chains, whether through market incentives, voluntary codes or trade regulations, require locally based knowledge co‐production, in which multiple stakeholders from agriculture and the food industry can benefit from working with locally based researchers. Read the free Plain Language Summary for this article on the Journal blog.

Corporate Voluntary Codes of Conduct for Preventing Human Trafficking through Corporate Social Responsibility Approach : An Empirical Study

This study explores the effectiveness of Corporate Voluntary Codes of Conduct(CVCC) for preventing human trafficking through a Corporate Social Responsibility (CSR) approach. The research involved a diverse sample size of 543 participants, including corporate governance professionals, international justice authorities, human trafficking combating establishments, and non- governmental organizations (NGOs). Both descriptive and inferential statistics were employed to test hypotheses, utilizing a range of statistical methods such as, Mean, Standard Deviation, Coefficient of Variation, Skewness and Kutosis and t-test, ANOVA, Correlation, Multiple Regression, and Confirmatory Factor Analysis (CFA ) respectively. The study utilized a 5-point Likert scale to assess 125 factors reduced to 15 combined factors out 77 factors, and 48 factors become redundant by administering data reduction technique related to preventing human trafficking, encompassing Corporate Voluntary Codes of Conduct and CSR. The findings were employed to develop a comprehensive frameworks to hold corporations accountable for human rights violations. Human Trafficking Controlling and Prevention Model. The results highlight the significance of CVCC inconjuction with CSR initiatives in addressing human trafficking issues beyond the confines of corporate codes. This research contributes to the evolving discourse on corporate responsibility and its potential impact on preventing human trafficking, emphasizing the need for a holistic approach that extends beyond regulatory frameworks

A Voluntary Code for ESG Ratings – A Pro Tem Solution?

A Voluntary Code for ESG Ratings – A Pro Tem Solution?

НОРМАТИВНЕ РЕГУЛЮВАННЯ ТЕХНОЛОГІЙ ШІ ТА ВПЛИВ НА БІЗНЕС

The functioning of business today is unimaginable without the use of artificial intelligence (AI) tools. The issue of regulating AI technologies in Ukraine and the impact of such algorithms on the development of business structures is relevant. The impact of AI on the digitalization of business can be observed in the following: automation of routine tasks; the ability to process large amounts of data and provide analytical reports; providing clients with personalized offers; the ability to predict future market trends. In Ukraine, AI regulation procedures began in December 2020 with the approval of the «Concept of Artificial Intelligence Development». The next step in regulation was the presentation of the «Roadmap for AI Regulation» document by the Ministry of Digital Transformation in October 2023. The document aimed to help Ukrainian businesses prepare for the adoption of a law that would be analogous to the European Union's EU AI Act. The adoption of the law was preceded by the publication of the European Strategy for AI in 2018, from which the Recommendations for Trustworthy AI were formed in 2019, and the Assessment of AI Trustworthiness in 2020. In parallel, the first Coordinated Plan on AI was published in 2018 as a joint commitment by EU member states. In February 2020, the European Commission approved a White Paper on AI. In April 2021, the European Commission began work on a full-fledged Regulation, which was finalized in March 2024 with the adoption of the EU AI Act, approved by the EU Council on May 21. The latest government document is the White Paper on AI Regulation in Ukraine: Vision of the Ministry of Digital Transformation. The main goals of the White Paper are: supporting businesses to increase competitiveness, protecting human rights from possible violations, and adapting Ukrainian legislation to EU standards. The White Paper describes the following prerequisites for developing a BOTTOM-UP approach: balancing interests; prudence and timeliness in implementing regulation; challenges in AI regulation; European integration «Brussels effect»; service function and product orientation. The starting tools are: regulatory sandbox, methodology for assessing the impact of AI on human rights, AI labeling tools, soft law tools (voluntary codes of conduct, general and sectoral recommendations), creation of a responsible AI web portal, and the final version of the White Paper. The logical continuation of Phase I is Phase II BOTTOM-UP – the adoption of the law analogous to the EU AI Regulation and the implementation of mandatory regulation through implementation. The White Paper proposes a path for AI regulation in Ukraine.

Digital distortions and interpretive choices: A cartographic perspective on encoding regulation

Rules as Code (RaC), which encompasses the conversion of legal and regulatory rules into computer code, is gaining traction internationally. This article analyses ‘digital distortions’ in RaC, which refer to disconnects between regulation and code that arise from interpretive choices in the encoding process. We contend that Boaventura de Sousa Santos’ ‘symbolic cartography of law’ provides valuable concepts for understanding digital distortions in encoding regulation. Specifically, we argue that the cartographic concepts of scale, projection, symbolisation and orientation highlight distortions that can arise from choices involving the documentation to code, logics to follow, languages to use and coded outputs to present. We demonstrate how these distortions arose in our attempt to convert the ePayments Code, an Australian voluntary code of conduct for consumer electronic payment transactions, into machine-executable code. The article concludes by underscoring the importance of greater awareness of interpretive coding choices and their implications for diverse users of digitised regulation.

Podcasting and ethics: Independent podcast production in New Zealand

In New Zealand – like in the US and UK – independently produced podcasts fall outside of local media regulations. New Zealand’s media laws and regulatory bodies remain broadcast and legacy media-focussed, so podcast content that has not been previously broadcast is not regulated or otherwise overseen. In the absence of regulation, this study explores the ways nine independent podcast producers from New Zealand self-govern their content, as well as their motivations for doing so. It is an investigation of the ways ‘amateur’ content producers approach media ethics, and more broadly podcast production in practice. Not guided or bound by formal publishing or editorial responsibilities, and mostly with no formal media training, study participants demonstrate adherence to journalistic principles. They consider ethical and editorial quandaries as they arise during the production process, factoring in the needs and disposition of their audience. This process is informed by their worldview, as well as their perspectives and experiences as media consumers. Though these podcasters champion the ethos of independent podcasting, the content of their shows is not free from third party influence. These podcasters are also parents, partners, employees, and colleagues; life roles that inform the content of their show. Though they push back against podcasting being legislated, these podcasters see value in creating an informal set of guidelines or a voluntary code of practice for podcasting in New Zealand. This project contributes to ongoing explorations of independent podcasting and podcasting practice, focussing on what defines, motivates, and informs self-driven practitioners.

Complaints about Violations of Voluntary and Pharmaceutical Industry-Run Medicine Promotion Codes in Canada.

Although regulation of pharmaceutical promotion in Canada is covered under the Food & Drugs Act and Regulations, in practice it is regulated by two codes, one a code administered by the Pharmaceutical Advertising Advisory Board and a second one controlled by Innovative Medicines Canada, the lobby organization for the large majority of the Canadian and foreign-owned pharmaceutical companies operating in Canada. This study examines complaints about code violations between 2012 and 2021 and puts those complaints and their outcomes into the context of the stringency of the codes, their governance, monitoring, penalties for violations, and compliance by companies. It combines the findings from this analysis with international experience with industry-run codes and concludes that overall, the Canadian codes are ineffective in controlling promotion. Finally, it offers recommendations for how regulation can be improved and how doctors' reliance on information from pharmaceutical companies can be reduced.

The voluntary code of conduct for dolphin watching in Port Stephens, Australia: is self-regulation an effective management tool?

In the absence of guidelines or government regulation for a rapidly expanding industry, dolphin watching operators in Port Stephens, New South Wales, Australia, formulated and adopted a voluntary code of conduct in 1996. This code was designed to reduce perceived pressures on dolphins and was updated to conform to the Australian National Guidelines for Cetacean Observation when they were released in 2000. Compliance to this code of conduct was assessed in a shore-based survey over the austral summer of 2002/03. Operator compliance was generally high for: number of dolphin watching boats per dolphin school; time spent by individual operators with dolphins; method of approach to dolphins; and frequency of cruises conducted per day. However, operators did not discriminate between dolphin schools containing calves and those that did not (equating to a breach of the national guidelines) and three of nine regular operators committed most breaches of the code, particularly with regard to boat-handling around dolphins and frequency of cruises conducted per day. The code’s aim in reducing exposure of dolphins to boats was not achieved as dolphin schools were subject to consecutive approaches by numerous boats and interactions also involved boats to which the code did not apply. This voluntary code is thus of limited value without revision, education and enforcement. The inability of a voluntary code to manage the number of operators and other watercraft highlights the need for management alternatives that will increase compliance by all users of the waterways. Furthermore, widespread assessments of compliance are necessary, particularly where assessments of the effects of cetacean-based tourism are being conducted. To determine whether identified impacts are a result of inappropriate management strategies, or non-compliance with suitable management, requires that management strategies are tested while simultaneously testing or ensuring compliance.

Effectiveness of Regulatory Policies on Online/Digital/Internet-Mediated Alcohol Marketing: a Systematic Review

BackgroundThe rapid growth of social networking sites and video sharing platforms has created an opportunity for the alcohol industry to employ advanced advertising and marketing approaches to target their audiences, increasingly blurring the lines between commercial marketing and user-generated content, which poses a challenge for effective regulation.MethodsWe conducted a systematic search through three peer-reviewed journal databases (WoS, PubMed, Scopus). Studies were included if published in English, after 2004, and assessed statutory regulation or voluntary industry codes, enacted by an EU or nation’s governmental agency or private entity, and with the intent to restrict digital alcohol advertising. In addition, we conducted a manual search of gray literature.ResultsA total of 4690 records were identified. After duplicate removal and full-text assessment, 14 articles were examined. Our findings indicate that children and adolescents may often be exposed to alcohol advertisements on social media and websites due to industry’s self-regulatory age-affirmation systems being largely ineffective at preventing under-aged access. Cases of self-regulatory violations by the alcohol industry, and increasingly innovative ‘gray-area’ advertising approaches have also been noted. Additionally, research illustrates a lack of developed statutory restrictions of digital alcohol advertising and instead continued reliance on voluntary industry self-regulation.ConclusionsThere is a substantial need for further research to examine the effectiveness of digital alcohol advertising restrictions in social media, websites and image/video sharing platforms. Moreover, there is a necessity for countries to develop comprehensive statutory frameworks, which would effectively restrict and monitor rapidly advancing digital alcohol advertising practices on new digital media.

Digitalization and disinformation: Current governance practices of the European Union

The article focuses on the contemporary problems of the digital space. It considers the threat of disinformation as a systemic phenomenon and a characteristic of “information disorder” and the post-truth era. The study reflects on the main theoretical ideas about the relations between agents of informational influence, involving various stakeholders. It also assesses potential and existing measures to counter disinformation. The experience of the European Union reflects in this context a new, supranational approach to regulation of the digital sphere. This article aims to comprehensively reflect the evolution and the current state of the European Union’s activities in the field of countering disinformation and assess the strengths and weaknesses of its policy. The article considers such important changes in the activities of the European Union as the adoption of the new Digital Services Act, the application of media blocking practices, the revision of previous documents, as well as the transition from a self-regulatory practice to the practice of co-regulation. Strategies, action plans, and resolutions, as well as voluntary codes of practice, provide the basis for the analysis. They reflect the strategic vision, measures, and results of the work of various EU institutions. The study showed that currently the EU has reached a high level of interaction with digital platforms in the fight against disinformation, while the nature of those interactions has gradually changed. Despite the central role of platforms in the EU’s policy, it combines new practices of co-regulation with the development of a particular strategic narrative and cooperation with civil society. At the same time, the current policy reveals a number of negative trends, including the politicization of the problem in the context of geopolitical contradictions, the stigmatization of certain information influence sources, as well as a lack of attention to “non-political” disinformation. Additionally, there are problems in EU policy related to the inconsistency of particular acts, the lack of transparency of numerous measures, and the difficulty of monitoring and evaluating certain actions.

Social Media and Party Manifesto: A study on the Key Factor in Deciding Election Outcomes

In the current study, researchers will look at the function of social media in managing political campaigns. Social media was a major factor in the 2014 general elections when it came to influencing voter behaviour. Numerous social media-related firms entered the scene starting in 2014. Because of this, a significant influence on domestic elections has been witnessed as a result of this new informational medium. The researcher discusses the function, significance, and potential of social media in politics as the primary subject of the current work. Additionally, the research will look at how political parties use social media for campaign management throughout different elections and how it affects voter behaviour. The Facebook-Cambridge Analytica data scandal, the rise of social media in Indian politics, the voluntary code of ethics established by social media platforms for the 2019 General Election, and finally its implications for ethics, political communication, and security will all be covered.

Unhealthy food and beverage marketing to children on digital platforms in Aotearoa, New Zealand

BackgroundChildren’s exposure to unhealthy food and beverage marketing has a direct impact on their dietary preference for, and consumption of, unhealthy food and drinks. Most children spend time online, yet marketing restrictions for this medium have had slow uptake globally. A voluntary Children’s and Young People’s Advertising (CYPA) Code was implemented in Aotearoa, New Zealand (NZ) in 2017. This study explores the Code’s limitations in protecting children from harmful food and beverage marketing practices on digital platforms accessible to children.MethodsA cross-sectional content analysis of company websites (n = 64), Facebook pages (n = 32), and YouTube channels (n = 15) of the most popular food and beverage brands was conducted between 2019 and 2021 in NZ. Brands were selected based on market share, web traffic analysis and consumer engagement (Facebook page ‘Likes’ and YouTube page views). Analysis focused on volume and type of food posts/videos, level of consumer interaction, nutritional quality of foods pictured (based on two different nutrient profile models), and use of specific persuasive marketing techniques.ResultsEighty-one percent of websites (n = 52) featured marketing of unhealthy food and beverages. Thirty-five percent of websites featuring unhealthy food and beverages used promotional strategies positioning their products as ‘for kids’; a further 13% used ‘family-oriented’ messaging. Several websites featuring unhealthy products also had designated sections for children, ‘advergaming,’ or direct messaging to children. Eighty-five percent of all food and drink company Facebook posts and YouTube videos were classified as unhealthy. Twenty-eight percent of Facebook posts for unhealthy products featured persuasive promotional strategies, and 39% premium offers. Nearly 30% of YouTube videos for unhealthy food and beverages featured promotional strategies, and 13% premium offers. Ten percent of Facebook posts and 13% of YouTube videos of unhealthy food and beverages used marketing techniques specifically targeting children and young people.ConclusionsThe voluntary CYPA Code has been in effect since 2017, but the inherent limitations and loopholes in the Code mean companies continue to market unhealthy food and beverages in ways that appeal to children even if they have committed to the Code. Comprehensive and mandatory regulation would help protect children from exposure to harmful marketing.

Analyzing the nexus of green economy, clean and financial technology

The connection between the green economy, technology, and finance has recently become a popular topic for analyzing economic and policy matters. Financial technology can provide not only an opportunity to tap into new pools of private capital to finance green and sustainable projects through innovative financial instruments but also provide support to clean technologies through the adoption of voluntary sustainability codes of conduct. However, there is still a lack of clear scientific evidence in the literature about how the green economy interacts with these relevant indicators of sustainable finance. Thus, this paper examines the time-varying causal relationship between indexes of financial technology (FinTech), clean technology (CleanTech), and the green economy (GECON), by applying the novel method proposed by Shi et al. (2018, 2020) on daily data from June 15, 2012 to December 15, 2021. This study finds a higher volatility and causality running from GECON to CleanTech and FinTech for the entire period. Furthermore, the green economy Granger causes FinTech and CleanTech with very significant episodes, especially at the start of the COVID-19 pandemic. The robustness of the results was checked with a rolling window and recursive evolving techniques that overall confirm bidirectional causal relationships between green economy and technology variables. The findings imply that global initiatives to achieve low-carbon economies need to be complemented with the use of clean technologies in the production process and the continuous digitalization of financial sectors. The promotion of clean technology production by governments and the increased interest of investors in FinTech industries will stimulate green economic growth.

Content Analysis of the Corporate Social Responsibility Practices of 9 Major Cannabis Companies in Canada and the US

The cannabis industry has sought to normalize itself and expand its markets in the 21st century. One strategy used by companies to generate positive public relations is corporate social responsibility (CSR). It is critical to understand these efforts to influence the public and politicians given the risks of increased cannabis use. To analyze cannabis industry CSR behaviors, determine their characteristics, and compare their practices with those of the tobacco industry. This qualitative study of CSR activities conducted between January 1, 2012, and December 31, 2021, evaluated 9 of the 10 largest publicly traded cannabis companies in the US and Canada. Data were collected from August 1 to December 31, 2021. The 10th company was excluded because it engaged in cannabis-based pharmaceutical sales but not CSR. A systematic review of corporate websites and Nexis Uni was performed, resulting in collection of 153 news articles, press releases, and Web pages. Charitable and philanthropic actions were included. Themes were identified and interpreted using modified grounded theory. CSR activities and spending. Nine major cannabis companies in the US and Canada engaged in CSR activities that encouraged increased consumption and targeted marginalized communities. Companies claimed these activities would mitigate the harms of cannabis prohibition, promote diversity, expand access to medical cannabis, and support charitable causes. They developed educational programs, sustainability initiatives, and voluntary marketing codes and used strategies similar to those used by tobacco companies to recruit public interest organizations as allies. These findings suggest that cannabis companies developed CSR strategies comparable to those used by the tobacco industry to influence regulation, suggesting that cannabis companies should be included when addressing commercial determinants of health.

Canadian Arts and Culture: Gender and Race in Leadership

Representation in leadership is important not only because it often shapes the ways in which priorities are set and decisions are made but it also signals who belongs and shapes the aspirations and assumptions about what is possible. Extensive research across sectors has exposed the under-representation of women, racialized people, Indigenous peoples, persons with disabilities and members of the LGBTQ2S+ community in leadership roles in Canada. A range of initiatives have been introduced including legislation, voluntary codes, educational programs and procurement and funding priorities to improve representation. In Canada, the arts and cultural sector is significant not just because it employs more than 851,456 people and receives hundreds of millions of dollars in government funding but because it plays a profound role in shaping values and culture with impacts far beyond its boundaries. While in recent years, organizations have been under growing pressure from communities and funders to address diversity and inclusion, there is often a gap between statements and actions. This paper examines dimensions of diversity and inclusion in the largest and most influential arts and cultural organizations in Canada (n=125). First, it examines the extent to which they have expressed commitments or aspirations to diversity and inclusion. Then it examines the representation of women and racialized people in leadership roles. Finally, it recommends the elements of a comprehensive strategy, grounded in the critical ecological model, to promote change at the sectoral, organizational and individual level.

MNEs’ CSR Decoupling: How Host-Country Stakeholder Pressures Matter

The literature suggests that organizations tend to use corporate social responsibility (CSR) as a means to establish and enhance legitimacy, particularly when they operate internationally. Meanwhile, organizations that adopt voluntary CSR initiatives, such as the United Nations Global Compact (UNGC), may decouple symbolic CSR claims from actual CSR performance. Despite extensive research on organizational decoupling and the legitimation strategies of multinational enterprises (MNEs), we have limited understanding of how host-country conditions affect MNEs’ CSR decoupling. Building on the idea of decoupling in institutional theory and the research on corporate adoption of voluntary codes of conduct, this paper analyzes how host-country stakeholder pressures will discourage CSR decoupling of MNEs that adopt the UNGC. We also identify possible contingencies, host country press freedom, and institutional and normative distances between home and host countries, and investigate how they moderate MNEs’ responses to host-country stakeholder pressures. We test our theory using an international sample of the UNGC signatories during the period 2007-2018. Our main analyses rely on a dataset containing 4,359 observations for 305 listed firms from 36 home countries, with negative CSR events (i.e., corporate violations of the UNGC principles) taking place in 117 host countries. The results support our predictions.