The current stage of scientific and technological progress, defined as the Fourth Industrial Revolution, is characterised by a profound paradox. On the one hand, humanity possesses an unprecedented toolkit for solving global challenges – from precision genome editing to complex climate modelling. On the other hand, the power and speed of research breakthroughs significantly outpace the development of adequate social, legal and, most critically, ethical systems designed to guide their application. Traditional research ethics, formed in the 20th century and centred around the principles of protecting personal autonomy (informed consent), minimising harm and ensuring fairness, is demonstrating its conceptual and practical inadequacy in the face of new realities.

This crisis is systemic and manifests itself in several dimensions.

First, the objects and subjects of research are increasingly not isolated individuals but complex heterogeneous systems: digital networks, ecosystems, social groups whose behaviour is mediated by algorithms. Classic procedures, such as obtaining individual informed consent, lose effectiveness when working with big data, where data are constantly aggregated, recombined and used in ways that are unpredictable at the time of their initial collection.

Second, the traditional “laboratory sterility” of the research process is disappearing. Science and innovation unfold in an open, globalised environment, where cause‑and‑effect relationships between a discovery and its broad socio‑environmental consequences become non‑linear, stretched in time and space, which makes responsibility for outcomes diffuse and collective.

Third, a phenomenon of “ethical lacunae” emerges – the appearance of fundamentally new dilemmas for which no ready normative precedents exist. These include questions concerning the moral status of artificial intelligence (AI) [1, 2, 3] and neurointerfaces [4], the right to “neuroprivacy”, or the attribution of responsibility for decisions made by autonomous systems.

We expect that in the near future the pressure of these factors will reach critical intensity. AI [1], synthetic biology [5], quantum computing [5] and climate engineering technologies will move from the experimental stage to the phase of large‑scale implementation, carrying not only colossal potential but also globally significant risks. In this context, the future of research ethics appears not as a matter of gradual evolution of existing codes, but as a problem of fundamental transformation and reassembly of the very paradigm of scientific responsibility.

To carry out a prognostic analysis and conceptualise the contours of this emerging paradigm. We argue that in the near future the central imperative will be a transition from retrospective and reactive ethics, which records already committed violations, to a proactive, holistic and constructive ethics that is inherently embedded in the fabric of the research process. Such ethics will have to operate not only with binary categories of “right/wrong” with respect to a single experiment, but also to assess systemic sustainability, long‑term social acceptability and the wisdom of chosen technological trajectories.

The paper successively examines key drivers of change, formulates the principles of a new model – the “ethics of entanglement” – and proposes possible institutional forms for its implementation.

Key drivers of change and emerging challenges

Data‑centric science and artificial intelligence. The shift of the scientific cognition paradigm towards data‑centricity (data‑driven science) and the total integration of AI represent not merely a technological upgrade but a fundamental transformation of epistemology – the very process of knowledge generation. This process will likely reach maturity when scientific hypotheses are not only tested but also generated by algorithms, and experiments are massively conducted in simulation environments or on digital twins of physical systems. This new epistemological reality gives rise to a series of deep ethical ruptures [6].

The epistemological challenge appears most vividly in the problem of the “algorithmic black box”. The traditional foundations of research ethics – transparency, reproducibility, the possibility of expert verification of reasoning – are called into question by complex neural network models. When an algorithm, having analysed millions of medical images, reveals a diagnostic pattern unknown to clinicians, the dilemma arises of whether it is permissible to accept the result without understanding the causal links. The requirement of “explainable AI” (XAI) transforms from a technical trend into an ethical imperative for critically important areas such as medicine, jurisprudence or infrastructure management. Ethical norms will dictate the necessity of a “presumption of explainability”: if the output of an algorithm affects peopleʼs lives or shapes the body of scientific knowledge, its logic must be accessible to human interpretation [1].

The second key challenge relates to fairness and inherited systematic biases. Data are not raw material but a social artefact bearing the imprint of historical and systemic inequality. Algorithms trained on such data not only reproduce but also scale and legitimise these biases. The fight against algorithmic bias will, in the near future, move from the realm of technical fixes into the sphere of systemic research ethics. This will require the introduction of mandatory ethical auditing of datasets, the principle of “fairness‑by‑design” at the algorithmic architecture level, and the development of strict guidelines for working with “hereditary data” – historical archives collected without compliance with modern informed consent standards.

The third set of problems concerns confidentiality and the consent model. The classic model of one‑time informed consent, tied to a specific study, becomes an anachronism in the era of total analytics and multiple data reuse. A transition is expected towards dynamic and adaptive consent models that allow participants to flexibly set the parameters for the use of their data. The widespread adoption of privacy‑enhancing technologies, such as federated learning and differential privacy, will become the de facto standard. Simultaneously, a complex ethical‑methodological question will arise regarding the “right to digital oblivion” in science and its reconciliation with the requirements of research reproducibility.

Finally, data‑centricity and AI radically change agency and the distribution of responsibility. In the chain “researcher – algorithm – data – result”, traditional concepts of authorship and blame are blurred. Institutional ethics in the near future will have to formalise these relationships through the introduction of roles such as “algorithm curators”, the development of “responsible AI” systems with full decision tracking, and a clear legal definition of “meaningful human control” over autonomous research systems. Thus, ethics in the era of data‑centric science will require not adaptation but a conceptual reboot, shifting focus from the protection of individual subjects to the governance of entire data and algorithm ecosystems.

Convergence of NBICS technologies and the problem of hybridity. The convergence of nano‑ (N), bio‑ (B), information (I), cognitive (C) and social (S) sciences represents a qualitative leap leading to the emergence of fundamentally new objects and methods, and consequently, new ethical problems. In the future, the applied use of convergent technologies will sharpen the main challenge: the inadequacy of traditional, sector‑specific ethical frameworks for assessing systems that are themselves borderline and hybrid.

The first challenge is ontological. Technologies erase the boundaries between the categories “living/non‑living”, “natural/artificial”, “human/machine”. The development of brain‑computer interfaces, the cultivation of cerebral organoids (“mini‑brains”) [7] and the creation of cybernetic organisms call into question the very boundaries of the moral community. Ethics will have to develop new criteria for determining the moral status of such hybrids, going beyond the framework of classical bioethics. Similarly, human enhancement technologies will raise questions not only of safety but also of fair access, long‑term impact on human identity, and the prevention of a new kind of social inequality.

The second challenge relates to managing complexity and unpredictability. NBICS systems possess emergent properties [8] that cannot be predicted by analysing individual components. The risk profile of a convergent technology (e.g., nanoparticles delivering genetic material on an AI signal) is not equal to the sum of the risks of its constituents. The ethical review of the future will have to integrate methods of systemic dynamic modelling and scenario forecasting to assess such synergistic risks at the earliest stages of R&D. The problem of “dual use” acquires a hypertrophied character, requiring ethics committees to have competencies in anticipatory ethics, closely linked with biosecurity and cybersecurity experts.

Example of an NBICS system: A smart biodegradable implant for drug delivery: nanoparticles (N) carry a genetic drug (B); the dosage is controlled by a microcomputer with AI (I) that reads brain signals via a neurointerface (C); all this is regulated by social norms and personal data laws (S).

“Neuro‑avatar” system for people with disabilities: a neurointerface with AI (C) reads brain signals (I), decodes them into commands for a bionic prosthesis made of new materials (N) that is connected to nerve tissue (B); the implementation of such a system requires changes in social infrastructure and ethical norms (S).

Why is NBICS a challenge for research ethics?

Thus, the term NBICS denotes a new class of technologies where the physical, biological, digital and social are so intertwined that they require a fundamentally new approach – the “ethics of entanglement” – which views any technology as part of a single, complex network, not as an isolated invention.

The institutional response to these challenges should be to overcome existing fragmentation. Instead of isolated committees for animal research or AI, the creation of hybrid interdisciplinary ethics committees capable of conducting holistic project assessment will be required. The social sciences and humanities component (S) will play a key role, acting not as a source of restrictions but as an active co‑constructor of the technological trajectory through early public engagement and the involvement of science and technology studies experts.

Globalisation, inequality and the ethics of distribution. The globalisation of the scientific process, characterised by the growth of international collaborations and mobility, paradoxically coexists with the deepening of systemic inequality. In the coming decade, ethics will have to address macro‑ethical questions of justice, sovereignty and structural dependency in the global system of knowledge production.

Classical “scientific colonialism” associated with the export of physical samples is transforming into more sophisticated forms. “Data extractivism” – the collection of genomic, medical and environmental data in developing countries to create commercial products in Global North countries – will require the entrenchment of principles of data sovereignty and fair benefit‑sharing. “Predatory” collaborations, where scientists from peripheral institutions serve only as data collectors, must give way to truly partnership‑based models.

The gap in the development of national ethical regulatory systems creates the risk of “ethics dumping” – the transfer of controversial research to regions with weak oversight. Countering this requires strengthening global minimum standards and developing a system of international certification of ethics committees. The open science movement, paradoxically, may exacerbate digital inequality if scientists from poor regions lack access to the infrastructure needed to work with open resources. The answer should be the principle of inclusive openness, implying investment in digital infrastructure and competencies in Global South countries.

Finally, globalisation exposes the conflict between the liberal‑individualistic model of ethics (with its emphasis on autonomy) and other cultural value systems. Future ethics will have to abandon blind universalism in favour of dialogical pluralism, providing for mandatory cultural contextual analysis, the involvement of local advisory boards, and the recognition of the diversity of knowledge forms.

Open science and the transformation of communication. In the future, open science principles (Open Access, Open Data) may become the new normative basis for research, generating a set of ethical dilemmas at the intersection of transparency, quality and security.

The ideal of total transparency confronts a paradox: on the one hand, it is a powerful tool against falsification; on the other hand, it creates the risk of “hyper‑transparency” and “witch‑hunts”, where the possibility of unlimited data audit may suppress risky but innovative research and create a toxic environment. Ethical norms will have to protect the “space for failure” and set reasonable boundaries for post‑publication peer review. At the same time, a culture of “open error” should be formed, where a correct correction is considered a sign of integrity.

The FAIR principles (Findable, Accessible, Interoperable, Reusable) will become an ethical obligation of the researcher [9, 10]. Making low‑quality or deliberately “noisy” data openly available will be considered an ethical violation. A key requirement will be the contextualisation of data through exhaustive metadata and “data stories”.

The economic model of open access based on article processing charges (APCs) risks creating a new structural inequality, leading to the “silencing” of scientists from poor regions. The only ethically sustainable model is one that includes mechanisms for the global redistribution of the financial burden and the development of free‑for‑authors platforms – “Diamond Open Access”. Openness also increases dual‑use risks, requiring the introduction of systems for proactive screening of research for security purposes before its publication, as well as “open‑by‑default” and “secure‑by‑design” models.

Finally, open science undermines the traditional article‑based model of authorship. The development of practices for citing datasets and code, systems of digital identifiers for all research outputs, and the recognition of the “invisible” labour of data stewards and infrastructure developers will be required.

Future contours of the new ethical paradigm

From reactive to proactive and constructive ethics. The traditional model, being reactive and retrospective, will prove inadequate in the face of technologies with non‑linear and long‑term consequences. A transition is needed to proactive (risk‑anticipating) and constructive (embedded in the process of creation) ethics.

Proactivity will be implemented through the introduction of Ethical Impact Assessment (EIA) as a mandatory iterative process throughout the entire project lifecycle, analogous to environmental impact assessment, and through the development of “Ethics‑shaping technologies” – algorithmic tools for early identification of ethical blind spots in research proposals.

Constructiveness means a transition to “ethics‑by‑design”, where ethical principles are materially embedded into technology architecture and research design [11]. This will require the appearance of a subsection “Ethical architecture of the study” in the methodological sections of articles and grant proposals.

Institutional preconditions for the transition will include the creation of “ethical technology sandboxes” for testing innovations in a real‑world context, reforming research evaluation systems to account for researchers’ “ethical capital”, and the emergence of new inter‑disciplinary professions – “ethical engineers” and “trust designers”. A key concept will be the management of “ethical debt” – the accumulated ethical problems arising from decisions that simplify development in the short term.

Ethics of entanglement. This paradigm proceeds from the ontological assumption of a deep interdependence of technological, biological, social and environmental systems that form a single dynamic network [12]. Its methodological core is a rejection of the search for linear causality in favour of an analysis of indirect and delayed effects (flow‑through effects) using system dynamics and network analysis, as well as the recognition of the ontological hybridity of the objects being assessed.

The ethics of entanglement radically expands the circle of moral patients, including ecosystems and non‑human actors, and asserts the principle of transgenerational responsibility, requiring long‑term scenario planning. Its institutional embodiments will be dynamic interdisciplinary committees and digital platforms for mapping ethical risks, visualising the entanglement network for a given technology.

The epistemology of responsibility within this paradigm becomes distributed (among all actors in the network) and anticipatory (focused on preventing potential harm rather than redressing harm already caused).

Expanding the concept of responsibility. In the near future, traditional responsibility as individual conscientiousness will become a necessary but insufficient minimum. It will evolve into a multidimensional construct.

A transition will occur from individual to collective and distributed responsibility, where institutions bear responsibility for creating the “ethical infrastructure”, and within projects, explicit matrices of the distribution of duties are developed. Responsibility will shift from retrospective (“Who is to blame?”) to prospective and anticipatory (“Who is obliged to prevent?”), including a duty of foresight and the management of ethical debt.

Responsibility will expand from narrowly professional to public and political, including responsibility for communication with society and for shaping the research agenda (agenda‑setting). Finally, the concept of responsibility‑as‑care will evolve, shifting focus from formal accountability to the attentive maintenance of the “health” of the research ecosystem and long‑term relationships.

Digitalisation of ethical oversight: opportunities and paradoxes. In the near future, digital tools will transform ethical oversight into an ecosystem of e‑oversight [13]. AI scanners will perform automated screening of publications and data for signs of misconduct, and dynamic monitoring systems will track the progress of research in real time. Blockchain technologies will provide immutable traceability of data and author contributions.

However, this transformation is fraught with paradoxes: the threat of total transparency and suppression of scientific freedom, the risk of shifting responsibility from humans to the system, digital inequality in access to oversight tools, and new threats to the confidentiality of the researchers themselves. The key principle should be the understanding that digital tools are meant to augment, not replace, human judgment and collective reflection. An ethical audit of the control algorithms themselves will be required, as will caution in the use of simplified metrics of “ethical integrity”.

Education and culture as the foundation of transformation. The success of the described changes will depend entirely on a deep transformation of education and professional culture. Ethics must be integrated into the training of scientists not as a separate course but as a cross‑cutting, context‑oriented line, developing “ethical imagination” and argumentation skills under conditions of uncertainty.

Cultural norms must shift from the hypercompetitive “publish or perish” to the values of reliability, openness and public good. This will require reforming assessment and reward systems, creating “ethics safe spaces” for discussing dilemmas, and forming a culture of “proactive transparency”. Training “ethically‑minded” leaders and supporting “ethical pioneers” will be critically important. Educational programmes will have to take into account the global cultural context and the pluralism of ethical traditions.

Institutional changes. Implementation of the new ethical paradigm will require a restructuring of the institutional landscape. Ethics committees will transform into interdisciplinary technology assessment councils with full public participation. Funding organisations will become architects of responsible agendas, introducing mandatory “ethical dossiers” and conditional financing mechanisms. Publishers will take on the role of curators of ethical standards for data and algorithms. New institutions will emerge: centres for ethical‑technology assessment (“ethics observatories”), international agencies for the accreditation of ethical standards, and offices of the scientific ombudsperson.

Universities and research centres will create internal “verticals of responsibility” and “sandboxes” for testing technologies. A key principle for the functioning of this ecosystem will be interoperability, ensured by end‑to‑end digital research identifiers, international agreements on mutual recognition, and standardised reporting formats.

The analysis carried out allows us to assert that the coming decade will mark not an evolutionary development but a paradigmatic shift in research ethics. The crisis of the classical model is a consequence of its fundamental mismatch with the realities of data‑centricity, the convergence of NBICS technologies, and globalised and open science. The ethics of the future must become an immanent, constituting element of the knowledge generation process itself.

The key vector will be the transition to proactive, constructive and entangled ethics, which implies a change in time perspective (from assessment to trajectory design), scale of analysis (from the experiment to the network of interdependencies) and agency of responsibility (from individual to distributed and accountable). Technological digitalisation will act simultaneously as a catalyst for this transformation and as an area of critical reflexive control. The culmination and necessary condition for change will be a deep transformation of scientific culture and education, shaping the scientist as an “ethical bilingual”.

Thus, in the coming decade, the contours of a new social contract between science and society must be formed, based not on blind faith in progress but on the demonstrated ability of the scientific community to anticipate risk management, ensure fairness and transparency. Science that wishes to retain public trust and its social licence will have to demonstrate not only intellectual power but also moral maturity. The future development of research ethics is a grand project of rethinking the very mission of science – a project on whose success depends whether the technologies of the coming decades become a source of new divisions or tools for building a more just, sustainable and humane future. Our children and grandchildren will see it.