Introduction
Resilient data and data systems are vital for effective crisis preparation and response. Providing, processing, and utilizing such data involve various stakeholders, including data experts, researchers, policymakers, and community members. In crisis management, accurate, timely, and reliable data is essential, as the urgency for accessing information can differ significantly from ordinary circumstances, especially during major events such as earthquakes or pandemics. Effective crisis management emphasizes eliminating barriers to data access, ensuring stakeholders can leverage open data. However, the drive to maximize benefits in crises may lead to attempts to acquire data without adequately respecting the privacy and confidentiality of individuals and communities. Such ethical oversights can erode trust in authorities and disrupt adherence to necessary measures during a crisis, complicating the equitable distribution of limited resources.
The UNESCO Recommendation on Open Science provides an internationally agreed definition, shared values, and guiding principles for open science. It also identifies actions conducive to a fair and equitable operationalization of open science for all at the individual, institutional, national, regional, and international levels (UNESCO, 2023; 2021). The values, principles, and objectives of the UNESCO recommendations pave the way to develop guidance on addressing ethical issues associated with data policy and open science in times of crisis.
This article explores the ethical issues surrounding data policy and open science management during crises. It highlights how these issues can be addressed through the framework established by UNESCO recommendations, which focus on good governance and a human-centered system approach. It aims to offer insights and practical tools to policymakers, scientists, frontline experts, and other stakeholders who provide, process, curate, and use data during crises. The discussion also emphasizes the importance of addressing ethical issues and considering ethical sensitivities while developing human-centered systems for sustainable digital technology. We hope the analysis framework could provide insight into lifecycle data management throughout crisis phases.
Ethical Concerns in Crisis Data Policies
Crisis data policies should be ready throughout crisis stages, such as prevention, preparedness, response, and recovery (RCO n.d.). Thus, different ethical concerns emerge. For example, for crisis preparedness, data sufficiency and effective flow of data matter most, followed by data integrity and reliability. However, such preparedness is insufficient for a swift and effective response unless better data resource allocation strategies are used to maximize efficiency and interoperability across domains and regions. Furthermore, data literacy will empower stakeholders to bridge data gaps by capacity building for crisis recovery. Preventing any crisis situation will involve effectively governance all the ethical concerns mentioned above (see more in Table 1). Therefore, the following crisis data analysis will embrace the ethical considerations we identify from the emergence stages.
Table 1
Ethical concerns for crisis data policies during emergency stages.
| EMERGENCY STAGES | ETHICAL CONCERNS IN CRISIS DATA POLICIES | |
|---|---|---|
| Prevention and | Preparedness |
|
| Response |
| |
| Recovery |
| |
Analysis of Ethical Considerations for Crisis Data Management
Ensuring a seamless, trustworthy, and interoperable flow of data
The first set of ethical problems consists of issues about the quality and quantity of data. In major and minor crises, maintaining sustainable data provision is problematic. The devastating effects of data scarcity have been experienced throughout history, with tragic examples such as Fukushima (International Science Council, 2024) and World War II. Recent crises, like the earthquakes in Turkey and Syria in 2023, showed us that data scarcity is not a historical event but a sustained problem in crisis management. One of the reasons for data scarcity is that data sources and the people responsible for data flow and management are usually affected by catastrophes. As a result, blockages in data flow can occur at a time when data is urgently needed. This problem is more magnified if the crises affect a larger geography, such as the COVID-19 pandemic. Even if data flow is sustained, there may be questions regarding the reliability and integrity of the available data. Data integrity may be breached, as exemplified in the COVID-19 pandemic. The scarcity of data and its possible flaws can manifest as a major problem in crisis management. Limitations in the interoperability of data are another issue that challenges crisis management. The lack of interoperability hinders the capability and effectiveness of various actors in using data. Addressing these problem areas related to access to data is the primary step in creating effective and responsible crisis management.
The core moral duty in crisis management is adeptly dealing with urgent situations that demand swift intervention to ensure the highest possible benefit and prevent irreparable harm for the majority. Consequently, providing a seamless, trustworthy, and interoperable flow of data stands as the primary ethical imperative for the ethics of crisis management.
Accurate, real-time, and continuous data flow is also examined within the scope of the right to information beyond its crucial role in crisis management. In its declaration titled ‘The Right to Information in Times of Crises’ (Mendel and Notess, 2020), UNESCO emphasizes that access to information during crisis periods is both an individual and public right. The UNESCO declaration frames public access to information as integral to achieving public health objectives and an essential part of crisis management and response rather than an external burden to authorities responsible for crisis management. Because the right to information is a fundamental human right, the authorities have a positive obligation to proactively disclose crucial emergency-related information encompassing health, budgetary, policy-making, procurement, economic, and benefits-related aspects. This obligation can only be fulfilled if data is made available.
From an ethical perspective, the right to information has both instrumental and inherent value. The instrumental value serves purposes such as providing benefits to the public and individuals, building trust in authorities, and complying with measures taken for crisis management. In addition to this instrumental value, it also serves an inherent value, such as preserving the respect for the autonomy of individuals and the public. Since access to accurate, comprehensive, and timely information is one of the key elements of autonomous decision-making, recognizing the right to information becomes an ipso facto element for exercising autonomy.
On the other hand, the right to information is not absolute. It is limited by other rights and values, such as the right to privacy, also articulated as a fundamental right in international legislation. Protecting data privacy and confidentiality are also ethical obligations for authorities since private data is a core element of human dignity, and violations of privacy and confidentiality of sensitive data pose significant risks to human integrity. The tension between disclosing information and breaching privacy and confidentiality is hard and should be addressed on a case basis to find and maintain a balance. The other concepts that raise this tension are public security, community surveillance, providing benefits, and enhancing the utilization capacity of measures taken by the authorities.
The balance in this multi-component strain may be achieved by establishing trust in individuals and the public toward the authorities. Establishing this trust is a crucial component for societies and individuals to adapt to crisis management, and it requires an organizational and executive structure that is in line with good governance principles. Placing good governance principles at the heart of crisis preparedness and response also serves sustainability in science and order in society during crises and recovery.
Communities and individuals should know that the resources will not be enough in a crisis to meet all needs. They should also be aware that if sacrifices are needed to allocate scarce resources, the decision-making process will be science-based, inclusive, collective, efficient, and transparent (United Nations, 2015). This system will be based on good governance principles and a delicate balance will be established to enable maximum benefit with minimum and temporary rights breaches.
Dilemma and reciprocal interaction between right bases approach and utilitarian approach
The second set of ethical issues arises from the utilitarian perspective, which advocates for the moral obligation to maximize benefits for the greatest number of people. Public health ethics, particularly crisis management ethics, typically align with a utilitarian approach. This alignment is reasonable given the pressing requirement to deliver benefits, such as timely and efficient rescue and emergency health services, to many people in need. Nevertheless, a sole utilitarian approach introduces an ethical dilemma by potentially sacrificing individual rights and freedoms in the pursuit of benefiting a significant number of people. Right to access adequate, reliable, and timely data, privacy and confidentiality, and limiting individual rights, as well as requiring some individuals to sacrifice some of their merits for the sake of others, are frequently overlooked values based on the maximum benefit argument.
In crisis management, the fundamental reason behind the demand for sacrificing certain rights and values to benefit the majority, or even considering this demand as an ethical obligation or responsibility, lies in the scarcity of resources. If scarce resources are not efficiently utilized, there is a high probability of irreversible damages, deaths, and losses occurring. During a crisis, one of the most constrained resources is time. There is very limited time available to make decisions quickly and accurately. The primary factor restricting time is the resilience periods of individuals or communities in coping with the devastation caused by the crisis. Access to accurate, real-time data is the key tool for effectively using time. Under normal circumstances, data ethics ensure the protection of data sources’ privacy, confidentiality, and informed consent, as well as safeguarding against negatives such as stigmatization and discrimination resulting from the use of personal data. However, the constraints during a crisis may lead those obligated to assist using data to overlook or even define these ethical responsibilities toward the data source as permissible neglect.
This situation compels individuals and communities facing a crisis to relinquish some rights and values they normally possess, obliging them to refrain from using these rights according to the decisions of those responsible for rescuing them and managing the crisis. Such a transfer of rights practically necessitates absolute trust in the authority’s structure, core values, and managerial skills to which the right is transferred.
The theoretical background of this transfer of rights is somewhat more complex. For a comprehensive discussion regarding the plausibility of limiting individual rights in crisis times, we have to delve deeper into three conceptual questions that also frame public health ethics. The first question pertains to the theoretical basis of the fundamental responsibility of authorities to protect or improve public health and well-being. The second question scrutinizes the foundations of crisis management policies’ goals and priorities. In contrast, the third question concerns the source, scope, and legitimacy of authority in implementing crisis management policies.
The first question can be rephrased as ‘what constitutes the basis of authority’s responsibility in crisis management?’ or ‘what is the foundation of the obligation to ensure public health and well-being in crisis management?’ Undoubtedly, as stated in the International Covenant on Economic, Social, and Cultural Rights1 and the World Health Organization’s Declaration of Alma Ata,2 this responsibility is based on the right to health. In countries that define health as a right at the level of both the public and individuals, authorities responsible for crisis management can be said to have the responsibility to carry out this management in the most accurate, timely, and equitable manner. This responsibility may grant the authority the right to temporarily suspend certain rights of individuals and communities.
Nevertheless, despite being declared in the International Covenant on Economic, Social, and Cultural Rights and the World Health Organization’s Declaration of Alma Ata, there is still no universal acceptance of health as a fundamental right. This situation diversifies the responses to the first question. In countries where the right to health is not recognized and health is positioned as a commodity subject to market conditions rather than defining it as a right, public health gives way to the notion of ‘community-level health.’ This concept corresponds to the collective well-being of a specific, limited community. The term ‘community’ in this context represents a much narrower and more variable group compared to the concept of the public. Since there is no right to health, there is no theoretical basis for limiting individual or community rights to benefit the public by the authority. This approach complies with the direct assignment of responsibility for the crisis management and preservation of well-being in crisis times to individuals or communities. This line of thinking justifies some bizarre exercises, such as purchasing emergency rescue services individually in the case of disasters. Note that such exercises breach the emergency procurement for crisis management and response strategies3 by creating an imbalance of fair allocation of scarce resources. In this situation, preserving public health, collective well-being, and fairness in accessing services may not be of significant importance. Moreover, there is no basis for sacrificing some rights to increase the benefit of the majority.
Another debate is whether public health is the sum of the health of individuals constituting the public or an entity with different norms and values than the sum of individual health conditions. The answer to this question varies in perspectives that do not view health as a fundamental right. The reductionist approach, which defines public health as the sum of the health of individual persons, solely defines health at the individual level, assigning all responsibility and authority for being and remaining healthy to the individual. According to this view, health authorities have very limited responsibility for public health. This responsibility includes services that can only be provided by public health authorities through financing, such as ensuring clean drinking water sources, constructing effective sewage systems, taking measures for air cleanliness, and supplying and providing access to vaccines.
On the other hand, the anti-reductionist perspective that sees public health as an entity based on societal values and norms holds health authorities responsible for preserving and improving public health. According to those who view health as a right, the foundation of this responsibility is the mentioned right. On the other hand, the anti-reductionist approach may also choose not to position health as a right. In this case, the concept of health for specific limited communities replaces the concept of public health, and the responsibility can be assigned to elements of social solidarity such as non-governmental organizations, religious belief groups, or cultural associations rather than managerial authorities.
The most notable difference between reductionist and anti-reductionist approaches emerges in their response to the third question: source, scope, and legitimacy of authority in implementing public health policies. According to the reductionist view, imposing restrictions on individual rights and freedoms by health authorities for public health is theoretically unfounded. The evaluation of the legitimacy of health authorities by the anti-reductionist approach can be either positive or negative. In other words, it can accept the restrictions imposed by health authorities as legitimate while holding them responsible for preserving and improving public health.
Although every theory does not universally accept the responsibility and authority of public health authorities for practices benefiting public health, the public’s involvement is necessary for the effectiveness and sustainability of most practices aimed at public health. This necessity is significantly higher in crisis management.
Mapping Open Science core value into crisis data ethics governance
Adopting open science in the design and development of strategies for ethical governance ensures the resilient arrangement of data ethics with the maximum benefits for the societal public. According to UNESCO (2021), the core value of Open Science extends the vision of certain works with five rules: ‘quality and integrity, collective benefits, equity and fairness, diversity and inclusiveness, and flexibility’. So, the way of crisis data ethics design and development should map these rules into real practices. Our analysis in the previous chapter shows four highlights that serve as catalysts for such work.
Right-based rules. Data ethics in crisis is to guide the most efficient and beneficial work of disaster mitigation supported by data. The right-based design of such data ethics means that it has been born in the current social culture and is rooted in local legislation and regional treaties. It will facilitate the implementation of such data works and accelerate data ethics governance to tackle hazardous situations. The ‘collective benefits, equity and fairness,’ as well as other Open Science core values, will surely be addressed in the right-based rules.
Human-centered approach. There are always gray areas where laws and policies are in progress, and chaos usually occurs in crises. Thus, beyond right-based rules, the based rules for data ethics policy design should insist on a human-centered approach. Less represented groups, such as young and old, women and children, indigenous and migrants, as well as others (United Nations, 2015), should always be encouraged and supported to join the ethical data dialogue to ensure truly ‘equity and fairness’ but also the ‘diversity and inclusiveness’ rules.
Interoperable data infrastructures. Crisis data ethics characterize lifelong data management. In governing crisis data ethics, the supporting infrastructures, either physical or virtual, should be ready to promote Open-science guided reliable data work utilizing interoperability at policy, technological, semantic, and community levels. Standards and protocols based on mutual trust should be prepared to ensure the quality and integrity of data. Commonly agreed vocabularies and instructions provide a semantic understanding of data entities. Hardware and software development under openly shared strategies to ensure fluent communication across systems and smooth flow of sundry data will enhance the performance of data infrastructures, thus guaranteeing resilient crisis data works.
Trustworthy research data ecosystems. Building upon the three principles, a trustworthy data ecosystem should be nourished to ensure sustained crisis data management in the long run. Such a trusted ecosystem pinpointing an open, collaborative environment and culture gains benefits from the ‘diversity and inclusiveness’ and, in turn, leads to flexibility in long-term crisis management. The mapping matrix of Open Science and crisis data ethics governance is shown in Table 2, with gray grids representing certain mapping relationships.
Table 2
Matrix of Open Science core values & crisis data ethics mapping.
| OPEN SCIENCE CORE VALUES | PRAGMATIC WAYS TOWARD CRISIS DATA ETHICS GOVERNANCE | |||
|---|---|---|---|---|
| RIGHT BASED RULES | HUMAN-CENTERED APPROACH | INTEROPERABLE DATA INFRASTRUCTURE | TRUSTWORTHY ECOSYSTEM | |
| Quality and integrity | ||||
| Collective benefit | ||||
| Equity and fairness | ||||
| Diversity and inclusiveness | ||||
| Flexibility | ||||
Implementing Crisis Data Ethics in AI Scenarios
Ethical considerations posed by Artificial Intelligence in crisis management. For various purposes, AI is used in crisis preparedness, management, and rehabilitation. In the preparedness stage, it can assist in activities requiring foresight, such as analyzing existing data for planning, developing intervention strategies, estimating the most effective, efficient, and rapid intervention methods through alternative simulations, determining potential resource requirements, and identifying how missing resources can be efficiently and timely procured. Using technologies like digital twins in conjunction with AI infrastructure can yield highly efficient results in simulating preparedness plans and developing alternative plans. During a crisis, especially when data flow is restricted or the reliability of incoming data is in question, AI can make assumptions based on previously collected and stored data. In cases where data is more reliable and interoperable, it can assist in making multifaceted decisions by continuously and accurately analyzing data of different qualities and contents from various sources. The decision-making process can simulate possible outcomes, achieving the most efficient result. Additionally, crisis management can assess the positive or negative mutual interactions of steps taken in different areas, considering the cross-sectoral effects of decisions. In rehabilitation, AI can conduct data analysis and simulations to detect damage, compensate for damages, and assist in rehabilitation with the least possible long-term and short-term losses (Table 3).
Table 3
Data actions as per artificial intelligence adoption in emergencies.
| EMERGENCY STAGES | USE OF ARTIFICIAL INTELLIGENCE |
|---|---|
| Preparedness |
|
| Response |
|
| Recovery |
|
| Prevention |
|
The benefits of using traditional AI in crisis management are enhanced and diversified by generative artificial intelligence since the generative abilities indicate the creation of new approaches and solutions rather than improving the effectiveness of the existing strategies. Generative AI can be used in a wide spectrum of tasks. Scenario generation is one of the areas in which generative AI can be useful. Creating realistic and diverse crisis scenarios based on historical or synthesized data can help train and test crisis management systems with various potential challenges, ensuring that preparedness plans are comprehensive and adaptable. Another area of use is offering dynamically adjusted scenarios based on real-time inputs, allowing for more adaptive training and testing and reflecting a crisis’s evolving nature. Generating alternative courses of action and their potential outcomes can enhance decision-making capacity, help authorities explore a wider range of choices, and anticipate the consequences of different decisions during the crisis.
Despite these potential benefits, there are various ethical concerns regarding the vast use of AI as a sole tool or decision support system in crisis preparedness, response, and rehabilitation. These concerns are even more amplified with the emergence of generative artificial intelligence because of the aforementioned data integrity and scarcity problems. Therefore, it is essential to consider ethical concerns, potential biases, and the need for careful oversight, especially in critical situations. Before embedding generative AI in crisis management, a broad awareness of the ethical problems it may impose or amplify should be considered.
These ethical concerns consist of three different sets of problems. The first problem set is deepening existing ethical issues in crisis preparedness and management through AI. Ethical considerations within this domain encompass the assurance of data flow, the preservation of data integrity, avoiding bias, the safeguarding of privacy, confidentiality, and intellectual property rights during data transmission, and taking into account individual rights and values that are overlooked in the pursuit of maximizing efficiency and benefit, are mentioned. The accuracy and validity of AI outputs are directly influenced by data integrity, interoperability, and continuity. In other words, the accuracy of AI is contingent upon the validity and reliability of the data it is fed. Consequently, the pivotal role of data in crisis management significantly influences the multidimensional ethical challenges posed by using AI.
The second problem arises from generative AI’s content generation, strategy development, planning, and multidimensional decision-making capabilities. While the foundation of this set still revolves around data integrity and continuous data flow, it is closely tied to the level of generative AI’s reinforced learning capabilities and the weight it places on humanitarian and ethical values in decision-making algorithms. As explained, the decision-making process of crisis management is oriented toward maximizing utility. How utility is defined or measured defines this process. Measuring utility in a disaster situation solely with quantitative indicators may result in the exclusion of some individuals or groups, leading to profound issues of equality, fairness, and justice. Therefore, it is plausible that the exclusive focus on maximizing quantitative utility in decision mechanisms in crisis management could potentially lead to the neglect of some disadvantaged groups, and how this risk is addressed in the AI algorithm should be considered.
Another issue with the decision-making process of generative AI is the black-box problem. Since the weights of various indicators in the decision process of the generative AI system are not known, the use of generative AI carries the risk of deepening the transparency and trust issues in crisis management. Trust in the authority is fundamental for public compliance with critical crisis decisions. It is rooted in the belief that the authority operates a decision-making process that adheres to good governance, fairness, equity, and transparency principles. The black-box problem inherent to generative AI may damage trust and impair social compatibility. Using Generative AI without improving it with ethics in design and ethics by design principles that cover all of its life cycles will intensify all these dilemmas. To fortify the feeling of justice, fairness, and trust in society in crisis management, the authorities should be transparent about full disclosure regarding where and how AI is used and what level of human oversight equipped with good governance tools is maintained for this use.
The third set of ethical issues emerges from the digital gap among countries or populations with access to Generative AI and limited or no means to incorporate this technology into their systems. It is a fact that there is already a gap that plays against developing countries in terms of technology and digitalization, and the vast use of AI where there is access to it will deepen the inequality between them. Considering the significant role of data and open science in crisis management, this fact should catalyze supporting those with limited access to digitalization to foster the required infrastructure and systems approach for effective crisis preparedness and response in compliance with UNESCO recommendations.
Conclusions and Future Envision
Drawing on the UNESCO Recommendation on Open Science (2021) and various legislative guidelines surrounding data ethics, this paper presents a comparative analysis of ethical data governance during crises. Key ethical data issues in the stages of emergency prevention, preparedness, response, and recovery include challenges related to data scarcity and ensuring an effective data flow, maintaining data integrity and reliability, and optimizing data resource allocation for enhanced efficiency, interoperability, and data literacy. Further analysis identifies three pathways toward effective ethical crisis data management: establishing a seamless, trustworthy, and interoperable data flow, fostering reciprocal interaction between rights-based and utilitarian approaches, and implementing Open Science mapping. In the context of AI scenarios, the analysis highlights ethical preparedness, the black-box problem, and digital gaps, emphasizing accessibility and usability as three critical challenges in designing and developing ethical data policies.
Looking ahead, the fundamental principles of UNESCO’s open science framework, alongside the human rights agenda, offer a solid foundation for addressing ethical concerns in crisis preparedness and response. Building an ethical framework for crisis preparedness and response systems that align with UNESCO’s open science principles and human rights entails the establishment of effective governance. This governance should uphold data integrity, respect human rights, and preserve human dignity while prioritizing scientific evidence obtained from reliable data, all under the principles of responsible research conduct. Additionally, it should emphasize transparency, community engagement, and the reinforcement of justice and trust in authorities.
Notes
[1] International Covenant on Economic, Social and Cultural Rights United Nations, 1996.
[2] Declaration of Alma Ata, World Health Organization, 1978.
[3] Emergency Procurement for Recovery and Reconstruction World Bank, 2015.
Funding Information
Zhang L. is supported by the National Natural Science Foundation of China (No. 72104229).
Competing Interests
Zhang L. is a member of the editorial board for the Data Science Journal.
Author Contributions
Perihan Elif Ekmekci designed the research framework and drafted the paper with Lili Zhang and Francis P. Crawley. All co-authors reviewed and finalized the paper.