Doris Schroeder

The PREPARED Code is a risk-based, values-driven framework that integrates research ethics and research integrity and is designed for a global audience. Developed over two years, this ambitious initiative required a collaborative, multidisciplinary effort led by an international team. The PREPARED team employed a range of methods to develop the code, including literature searches, scoping reviews, empirical studies, targeted consultations, ethical and legal analyses, and public consultation. This chapter explores the processes and methods used to develop the PREPARED Code, highlighting how real-world challenges in research ethics and research integrity during crises were identified, analysed and validated by stakeholders. It describes how these challenges were aligned with universally recognised moral values and grouped as risks, and how the risks were transformed into a clear, focused and jargon-free code of conduct. It also details the final stages of development, which involved iterative refinement of the code from Version 1 to Version 13, through extensive consultation and review.

ABSTRACT Benefit sharing aims to achieve an equitable exchange between the granting of access to a genetic resource and the provision of compensation. The Convention on Biological Diversity (CBD), adopted at the 1992 Earth Summit in Rio de Janeiro, is the only international legal instrument setting out obligations for sharing the benefits derived from the use of biodiversity. The CBD excludes human genetic resources from its scope, however, this article considers whether it should be expanded to include those resources, so as to enable research subjects to claim a share of the benefits to be negotiated on a case‐by‐case basis. Our conclusion on this question is: ‘No, the CBD should not be expanded to include human genetic resources.’ There are essential differences between human and non‐human genetic resources, and, in the context of research on humans, an essentially fair exchange model is already available between the health care industry and research subjects. Those who contribute to research should receive benefits in the form of accessible new health care products and services, suitable for local health needs and linked to economic prosperity (e.g. jobs). When this exchange model does not apply, as is often the case in developing countries, individually negotiated benefit sharing agreements between researchers and research subjects should not be used as ‘window dressing’. Instead, national governments should focus their finances on the best economic investment they could make; the investment in population health and health research as outlined by the World Health Organization’s Commission on Macroeconomics and Health; whilst international barriers to such spending need to be removed.

Codes of ethics and conduct typically demand the highest standard of ethical behaviour from every single employee. This implies a democratic or lobbyist understanding of ethics in business. The contrasting view would argue that business ethics is an elitist undertaking that can only be instigated from the top, by managing directors or owner managers. This article looks at three types of ethical businesses, three types of approaches to ethical problem‐solving, and three possible incentives for ethical business to see which of the above is more convincing. It will be argued that the impulse for value‐driven business has to come from the top. Only top management and owner managers have the opportunity to foster major change in business practice.

This chapter explains the conceptual foundations of the PREPARED Code, which together provide the credibility required to justify adding yet another ethics code to the thousands that already exist. The code is built on real-world risks identified in nine languages rather than, for instance, on drafters’ expertise, thereby making it as precisely honed an instrument as possible to cope with the real-world ethics and integrity challenges experienced during a pandemic. The code is values-driven, focused on the values of fairness, respect, care and honesty, to harness the motivational power of moral values and to provide an easily understandable, globally applicable moral framework. Unlike most other ethics codes, the PREPARED Code unites research ethics and research integrity guidance into one, to ensure that a culture of integrity rather than a box-ticking mentality is fostered. The short, jargon-free code text addresses all research disciplines and, most importantly, it is based on extensive input from a wide range of stakeholders, including highly marginalised populations, to ensure that it is fit for purpose.

Research ethics and integrity codes lay the foundation for ethical research. However, stakeholders in research may struggle to move from reading codes to applying them, especially during times of crisis when there is increased uncertainty and risk. To bridge this gap, the PREPARED project team has developed adaptable tools to support implementation of the PREPARED Code for research ethics and research integrity during pandemics. These include training clips to accompany each code article, an experiential learning app, and how-to style guidelines for enhancing the resilience of stakeholder-specific processes. In this chapter, we summarise PREPARED’s approach to code implementation, elaborate on each tool developed, and provide tips for future initiatives seeking to improve the practical application of ethics codes. We also present an example of how our tools were utilised during an African Vaccine Regulatory Forum (AVAREF) training session and provide a resource bank to support the integration of our materials into ethics training programmes.

What do ethics codes and guidelines tell us about who is vulnerable in research? To what are they vulnerable? And how might this vulnerability be addressed? These questions guided our analysis of 57 research ethics codes and guidelines that mention the involvement of vulnerable persons in research. The chapter draws upon the findings from this analysis to help explain how and why some people might be excluded from research unnecessarily. The investigation is also informed by the findings from an empirical study into the perspectives of 22 researchers who have extensive experience of undertaking research with vulnerable individuals. Their encounters with research governance measures, including research ethics approvalEthics approval and research ethics codes, expose challenges that can exacerbate unnecessary exclusionExclusion from research. We learn from the researchers and the research ethics codes that governance measures intended to protect the vulnerable from exploitationExploitation in research can lead to unfair exclusionExclusion when applied in a blanket fashion. Research governance guidelines and processes must encourage a more nuanced understanding of what it means to be vulnerable in research to ensure that people are not left behind unnecessarily.

Engaged researchEngaged research, which strengthens research teams through community researchersCommunity researchers, offers many opportunities and challenges. From better access to community members who are hard to reach, to the collection of more meaningful and authenticAuthentic data, and greater trustworthiness of research findings, the benefitsBenefits for research are manifold. However, research has also shown that community researchersCommunity researchers might be overtly biased, only collect superficial data or lack the confidenceConfidence to probe deeply enough, among other challenges. Simultaneously, the literature on community researchersCommunity researchers is heavily biased towards high-income countries, and there is very little to be found on experiences from low-and middle-income countries that goes beyond assistance in obtaining informed consentInformed consent. This chapter starts to close that gap by presenting a case study involving 12 community researchersCommunity researchers from the South African San community. Collecting no personal dataPersonal data and obtaining all research input through community researchersCommunity researchers shows that research led by vulnerable groups for vulnerable groups is possible. It is one way of ensuring that the San, and wider research communities, have access to research they can trustTrust.

The Indigenous San peoples, often referred to as South Africa’s “First Peoples”, experienced a violent history of displacement and genocideGenocide. Modern-day San still suffer from the intergenerational traumaIntergenerational trauma inflicted by colonists as well as discriminationDiscrimination, marginalisationMarginalisation and impoverishment. In addition, the South African San are collectively labelled as a vulnerable group, whose inclusionInclusion in research should be reduced to a minimum if one follows traditional protection mechanisms from research ethics. The purpose of this chapter is to let the San define “vulnerability” for themselves through workshops and community-administered surveysSurvey. It became clear that San representatives did not perceive the word “vulnerable” to be pejorative per se. On the contrary, it served as a useful umbrella term to cover a vast range of issues and problems. However, it also became clear that the term is perceived as negative when external actors block research—in a potentially patronisingPatronising manner—that could be beneficial to the community. Equitable research partnershipsEquitable partnerships between researchers and South African San community members, facilitated by the South African San CouncilSouth African San Council and guided by the San Code of Research EthicsSan code of research ethics, can lead to rich data that have the potential to assist the community in finding ways out of vulnerability. Externally labellingLabelling them as “vulnerable” as a group and blocking access to research cannot be the solution.

Sex work is one of the most stigmatised professions in many parts of the world. In Kenya, where it is also illegalIllegal, sex workers can even face rapeRape and abuseAbuse at the hands of law enforcement agents when it becomes known how they earn a living. As a result, sex workers rarely disclose their profession to familyFamily members, let alone outsiders. This means that the involvement of Kenyan sex workers in research over the years has been highly risky, as most research efforts collect personal dataPersonal data. This chapter describes a four-stage investigation into how sex workers define vulnerability for themselves. The approach was community-driven, community-led and, in the main, community-analysed. Importantly, no personal dataPersonal data was collected, and all engagement was undertaken by trusted facilitators and a sex worker community researcherCommunity researchers. Findings showed that structural vulnerabilities are intersectionalIntersectional and stigma-aggravated. At the root may be povertyPoverty, but the stress created by being stigmatised can lead to mental healthMental health problems which, in turn, can aggravate povertyPoverty and stigmaStigma. The KiswahiliKiswahili term mnyongeMnyonge may capture the essence of vulnerability, expressing a sense of weaknessWeakness without the ability to defend oneself.

Leaving no one behindLeaving no one behind is the main transformative promise of the UN’s 2030 Agenda for Sustainable Development. It encapsulates the 21st-century mission of inclusionInclusion. This chapter introduces the main mission of this book: leaving no one behindLeaving no one behind in research. It provides the context for all the chapters that follow by explaining what it means to leave no one behind in research, how the protection-inclusionInclusion dilemma for vulnerable groups and individuals is generated, and how risksRisk and mistrust in research might be reduced. The book was written in collaboration with two groups who are traditionally labelled as highly vulnerable and are therefore often excluded from research: the South African San community and an impoverished sex worker community in Nairobi. Working closely throughout all research stages with the two communities, including co-authorship of this book, we prioritised research-participant needs over researcher needs, aiming for minimally risky and minimally burdensome research, as well as increased trustTrust in researchers. This involved foregoing the collection of personal dataPersonal data and obtaining all research input through community researchersCommunity researchers. It led to a potential alternative to exclusionExclusion from research, namely research led by vulnerable groups for vulnerable groups.

This concluding chapter summarises how we challenged the protection-inclusionInclusion dilemma in order to avoid leaving people behind in research unnecessarily. The fact remains that individuals from highly impoverished, stigmatised groups in lower-income settings face a high likelihood of being harmed and exploited in research. However, excluding them from research is not the answer and can also be seen as a patronisingPatronising interference in the lives of people who might benefitBenefits from research and might welcome long-term relationshipsLong-term relationships of trustTrust with researchers. Achieving long-term relationshipsLong-term relationships of trustTrust requires research methods that strive to be minimally burdensome and minimally risky. In the context of working with the South African San and the Nairobi sex workers, this means that we had to avoid all personal dataPersonal data collection. It also required community involvement at all stages of the research. A key consideration, often lacking from the wider literature, is how community researchersCommunity researchers are identified, trained and supported. In our work, this involved trusted community members leading the trainingTraining, and the use of storytelling, humourHumour and SocraticSocratic methods to encourage reflexivityReflexivity, compassionCompassion and sensitivity. Notable findings generated from our work include that ‘vulnerability’ is not lost in translation, even though the English term is not readily translatable into relevant other languages, or where the main stressors—such as the stigmatisationStigmatisation, discriminationDiscrimination, marginalisationMarginalisation and abuseAbuse besetting the sex worker community—cannot simply be subsumed under “vulnerability”, because they involve specific individual actors who bring about hardship. Our work identifies an approach for leaving no one behindLeaving no one behind in research that can stimulate a deeper understanding of how equitable research partnershipsEquitable partnerships with highly marginalised groups can be achieved.

PrivacyPrivacy and data protection are concerns raised about most digital technologies. The advance of artificial intelligenceArtificial intelligence (AI) has given even higher levels of prominence to these concerns. Three cases are presented as examples to highlight the way in which AIArtificial intelligence can affect or exacerbate privacyPrivacy concerns. The first deals with the use of private data in authoritarian regimes. The second looks at the implications of AI use of genetic data. The third concerns problems linked to biometric surveillance. Then follows a description of how privacy concerns are currently addressed via data protection regulationRegulation and a discussion of where AI may raise new challenges to existing data protection regimes. Current European data protection law requires data protection impact assessmentImpact assessment. This chapter suggests that a broader AI impact assessmentImpact assessment could broaden the remit of such an assessment to offer more comprehensive coverage of possible privacyPrivacy concerns linked to AIArtificial intelligence.

Surveillance capitalismSurveillance capitalism hinges on the appropriationAppropriationand commercialisationCommercialisationof personal dataPersonal datafor profitProfit-making. This chapter spotlights three cases connected to surveillance capitalismSurveillance capitalism: data appropriationAppropriation, monetisationMonetisationof health dataHealth data and the unfair commercial practice when “free” isn’t “free”. It discusses related ethical concerns of powerPowerinequalityInequality, privacyPrivacy and data protection, and lack of transparencyTransparencyand explainabilityExplainability. The chapter identifies responses to address concerns about surveillance capitalismSurveillance capitalism and discusses three key responses put forward in policy and academic literature and advocated for their impact and implementation potential in the current socio-economicSocio-economicsystem: antitrustAntitrustregulationRegulation, data sharingData sharingand accessAccess, and strengthening of data ownershipData ownership claims of consumersConsumer/individualsIndividual. A combination of active, working governance measures is required to stem the growth and ill-effects of surveillance capitalismSurveillance capitalism and protect democracyDemocracy.

The concluding chapter highlights broader lessons that can be learned from the artificial intelligenceArtificial intelligence (AI) cases discussed in the book. It underlines the fact that, in many cases, it is not so much the technology itself that is the root cause of ethical concerns but the way it is applied in practice and its reliability. In addition, many of the cases do not differ radically from ethics cases related to other novel technologies, even though the use of AIArtificial intelligence can exacerbate existing concerns. Ethical issues can rarely be resolved to everybody’s full satisfaction, not least because they often involve the balancing of competing goods. What is essential is space for human reflection and decision-making within the use of AI. Questions about what we can and should do, why we should act in particular ways and how we evaluate the ethical quality of our actions and their outcomes are part of what it means to be human. Even though Immanuel KantKant, Immanuel believed that a good will is the only thing in the world that is ethical per se, a good will alone does not suffice where complex consequences may not be obvious. The complex nature of AIArtificial intelligence systems and their interaction with their human, social and natural environment require constant vigilance and human input.

The concern that artificial intelligenceArtificial intelligence (AI) can be used to manipulate individualsIndividual, with undesirable consequences for the manipulated individualIndividual as well as society as a whole, plays a key role in the debate on the ethics of AI. This chapter uses the case of the political manipulationManipulation of voters and that of the manipulationManipulation of vulnerable consumersConsumer as studies to explore how AI can contribute to and facilitate manipulationManipulation and how such manipulationManipulation can be evaluated from an ethical perspective. The chapter presents some proposed ways of dealing with the ethics of manipulationManipulation with reference to data protection, privacyPrivacyand transparencyTransparency in the of use of data. ManipulationManipulation is thus an ethical issue of AIArtificial intelligence that is closely related to other issues discussed in this book.

There is much debate about theUnfair discrimination ways in which artificial intelligenceArtificial intelligence (AI) systems can include and perpetuate biases and lead to unfair and often illegalIllegal discrimination against individualsIndividual on the basis of protected characteristics, such as age, race, gender and disabilityDisabilities. This chapter describes three cases of such discriminationDiscrimination. It starts with an account of the use of AIArtificial intelligence in hiring decisions that led to discrimination based on gender. The second case explores the way in which AI can lead to discrimination when applied in law enforcementLaw enforcement. The final example looks at implications of bias in the detection of skin colour. The chapter then discusses why these cases are considered to be ethical issues and how this ethics debate relates to well-established legislation around discriminationDiscrimination. The chapter proposes two ways of raising awareness of possible discriminatory characteristics of AI systems and ways of dealing with them: AI impact assessmentsImpact assessment and ethics by design.

This chapter introduces the themes covered by the book. It provides an overview of the concept of artificial intelligenceArtificial intelligence (AI) and some of the technologies that have contributed to the current high level of visibility of AI. It explains why using case studiesCase study, case studies is a suitable approach to engage a broader audience with an interest in AI ethics. The chapter provides a brief overview of the structure and logic of the book by indicating the content of the cases covered in each section. It concludes by identifying the concept of ethics used in this book and how it is located in the broader discussion of ethics, human rightsHuman rightsand regulationRegulationof AIArtificial intelligence.

Dignity is a very prominent concept in human rightsHuman rights instruments, in particular constitutions. It is also a concept that has many critics, including those who argue that it is useless in ethical debates. How useful or not dignity can be in artificial intelligenceArtificial intelligence (AI) ethics discussions is the question of this chapter. Is it a conversation stopper, or can it help explain or even resolve some of the ethical dilemmas related to AI? The three cases in this chapter deal with groundless dismissal by an automated system, sex robotsSex robotsand care robotsCare robots. The conclusion argues that it makes perfect sense for human rightsHuman rights proponents to treat dignity as a prime value, which takes precedence over others in the case of extreme dignity violations such as torture, human trafficking, slaverySlavery and reproductive manipulationManipulation. However, in AIArtificial intelligence ethics debates, it is better seen as an equal among equals, so that the full spectrum of potential benefits and harmsHarm are considered for AI technologies using all relevant ethical values.

In 2015, 193 nations cameAI for Good together to agree Agenda 2030Agenda 2030: 17 goals ranging from the elimination of poverty to the building of partnerships to achieve those goals. The spirit of the UN Sustainable Development GoalsSustainable Development Goals (SDGs) (SDGs) is to leave no one behind. Artificial intelligenceArtificial intelligence (AI) has a great potential to assist in reaching the SDGs. For instance, using algorithmsAlgorithms on new and vast agricultural data sets can improve the efficiency of agriculture practices and thereby contribute to SDG 1, “Zero hunger”. However, the high energy consumption, computational resources and levels of expertise required for AIArtificial intelligence can exacerbate existing inequalitiesInequality. At the same time, potentially useful AI applications such as seasonal climate forecastingSeasonal climate forecasting have led to the accelerated laying off of workers in PeruPeru and credit denial to poor farmers in ZimbabweZimbabweand BrazilBrazil. If AI for GoodAI for Good is to be truly realised, AI’s potential to worsen inequalityInequality, to overexploit resources, to be undertaken through “helicopter researchHelicopter research” and to focus on SDGSustainable Development Goals (SDGs) issues relevant mainly to high-income countries must be overcome, ideally in close collaboration and engagement with potential beneficiaries in resource-limited settingsResource-limited settings.

Artificial intelligenceArtificial intelligence(AI) canRight to life, liberty and security of personsupport individualsIndividual’ enjoyment of lifeLife, libertyLibertyand securitySecurity, but it can also have adverse effects on them in a variety of ways. This chapter covers three cases affecting human lifeLife, liberty and security: one in transportationTransport(self-driving carsSelf-driving cars), one in the home (smart securitySecurity systems) and one in healthcareHealthcareservices (adversarial attacksAdversarial attack). The chapter discusses ethical questions and three potential solutions to address AIArtificial intelligencehuman rightsHuman rights issues related to lifeLife, liberty and securitySecurity of persons: defining and strengthening liabilityLiability regimes, implementing quality managementQuality management systems and adversarial robustnessAdversarial robustness. AI developersDevelopers, deployers and usersUsers must respect the sanctity of human lifeLife and embed, value and respect this principle in the design, development and use of their products and/or services. Critically, AIArtificial intelligence systems should not be programmed to kill or injure humans.

How can an ethics code achieve impact? The answer is twofold. First, through adoption by influential research funders, who then make it mandatory for their award recipients. This is the case with the Global Code of Conduct for Research in Resource-Poor Settings, which was adopted by both the European Commission and the European and Developing Countries Clinical Trials Partnership shortly after its launch in 2018. Second, an ethics code can achieve impact when researchers use it for guidance whether it is compulsory or not. This is most likely to happen with codes that were developed transparently with all research stakeholders involved. This chapter will outline how the GCC was developed, and in particular how external stakeholders were systematically engaged, how existing codes were carefully analysed and built upon, and who the early adopters were.

The UN’s 2030 Agenda for Sustainable Development calls for more research and innovation to end poverty, leaving no one behind – and yet the export of unethical practices from high-income to lower-income settings is still a major concern. Such ethics dumping occurs in all academic disciplines. When research is regarded, on the one hand, as a dirty word among vulnerable populations who face ethics dumping, and, on the other, as a solution to many of humanity’s problems, how can the resulting gulf be bridged? This book describes one initiative to counter ethics dumping: the development and promotion of the Global Code of Conduct for Research in Resource-Poor Settings.

The Global Code of Conduct for Research in Resource-Poor Settings claims global applicability and promotes respect as one of its four values. Hence, the code anticipates potentially unresolvable differences between cultures, while maintaining it is globally valid. Examining, but discarding, several possibilities to deal with normative relativism, this chapter argues, with Beauchamp and Childress (2013, Principles of Biomedical Ethics, 7th edn. Oxford University Press, New York) that values can be internal to morality itself, allowing their global applicability.

Ethics dumping occurs in collaborative international research when people, communities, animals and/or environments are exploited by researchers. Exploitation is made possible by serious poverty and extreme power differentials between researchers from high-income countries and research stakeholders from low- and middle-income countries (LMICs). To prevent its occurrence, the risks of exploitation have to be tackled. This chapter describes 88 risks identified for collaborative international research, categorized according to four values: fairness, respect, care and honesty. The risks were identified in a broad-based consultative exercise, which included more than 30 members and chairs of ethics committees in LMICs, representatives from vulnerable populations in LMICs, and an open call for case studies of exploitation. The findings of the exercise contributed to the development of the Global Code of Conduct for Research in Resource-Poor Settings.

An ethics code is not enough to avoid ethics dumping. Ethics codes can inspire, guide and raise awareness of ethical issues, but they cannot, on their own, guarantee ethical outcomes; this requires a multifaceted approach. For research in resource-poor settings, engagement is crucial. Such engagement has been built into the Global Code of Conduct for Research in Resource-Poor Settings as a requirement, but how can it be put into practice? An approach for ethical community engagement is presented in this chapter, which also includes suggestions for an accessible complaints mechanism. At the institutional level, we tackle the question of concluding fair research contracts when access to legal advice is limited. Throughout, at a broader level, we show how the four values of fairness, respect, care and honesty can be used to help guide decision-making and the practical application of the code.

Values inspire, motivate and engage people to discharge obligations or duties. This chapter defends the values approach in the context of guarding against ethics dumping, the practice of exporting unethical research from higher-income to lower-income settings. A number of essential questions will be answered: What are values? What is the meaning of the word “value”? Why does it make sense to choose values as an instrument to guide ethical action in preference to other possibilities? And what is meant by fairness, respect, care and honesty? It is concluded that values can provide excellent guidance and aspiration in the fight against ethics dumping, and are therefore a well-chosen structure for the Global Code of Conduct for Research in Resource-Poor Settings.