Earlier this summer (May of 2022) The committee on Responsible Computing Research of the National Academies of Sciences, Engineering and Medicine in the USA published a Consensus Study Report entitled Fostering Responsible Computing Research, Foundations and Practices (following a request from the National Science Foundation). I attended its virtual launch and duly downloaded the prepublication copy (a version that is subject to further editorial correction published by The National Academies Press) and read the document.

Contributions came from the following committees: Ethics and Governance of Computing Research and Its Applications; Computer Science and Telecommunications Board; Division on Engineering and Physical Sciences; National Academies of Sciences, Engineering, and Medicine.

Consensus Study Reports

Consensus Study Reports document the evidence-based consensus on the study’s statement of task by an authoring committee of experts, typically including findings, conclusions, and recommendations based on information gathered by the committee and the committee’s deliberations. Each report is subjected to independent peer-review and represents the position of the National Academies on the statement of task.

The Statement of Task is quite lengthy, but includes the following descriptions (my editing):
the study will explore ethics and governance issues associated with the personal and social consequences of computing research and its applications, considering such tasks as
1) Guiding principles, tools, and practical approaches for identifying and addressing ethical issues;
(2) The feasibility and likely performance of research governance frameworks and regulatory regimes, and related best practices that research funders, research-performing institutions, and individual researchers can leverage to formulate, conduct, and evaluate ethical research and associated activities;
(3) Multidisciplinary approaches to understanding ethical issues in computing research;
(4) How these approaches can empower the research community to develop and pursue socially productive practices; and
(5) Ways to promulgate ethical principles and responsible practices and sustain attention to them in the computing research community, including through education and training.

The statement also proposes a series of questions to be addressed:
(a) How do ethics and governance issues and needs present differently in different research contexts? Are there other ethics and governance issues that apply more broadly across many or most areas of computing research?
(b) What set of research governance frameworks or regulatory regimes are feasible in each of these contexts?
(c) How might research governance take place at different granularities and modalities of governance, such as community, organizational, local, regional, national, and international?
(d) What empirical evidence exists for how these research governance frameworks or regulatory regimes might correspond to ethically desirable outcomes?
(e) What is the current relative maturity level of ethics and governance concepts in different aspects of the computing research space? Which areas are the most advanced and can their relative maturity be leveraged into use elsewhere in computing?
(f) What incentives or contextual changes would be effective in helping computing researchers, and those who develop subsequent applications, place more emphasis on ethical considerations? For which existing, and likely future, stakeholders are such changes compatible with current incentives?

The Committee members predominantly came from US universities (with one representative from Australia), but also included representatives from Microsoft, Google and the Carnegie Endowment for International Peace.

Conclusions

The process led to the committee coming to three conclusions that underpin the report recommendations:

Conclusion 1. To be responsible, computing research needs to expand to include consideration of ethical and societal impact concerns and determination of effective ways to address them.

Conclusion 2. To be responsible, computing research needs to engage the full spectrum of stakeholders and deploy rigorous methodologies and frameworks that have proven effective for identifying the complicated social dynamics that are relevant to these ethical and societal impact concerns.

Conclusion 3. For computing technologies to be used responsibly, governments need to establish policies and regulations to protect against adverse ethical and societal impacts. Computing researchers can assist by revealing limitations of their research results and identifying possible adverse impacts and needs for government intervention.

Recommendations

Recommendation 1. The computing research community should reshape the ways computing research is formulated and undertaken to ensure that ethical and societal consequences are considered and addressed appropriately from the start.

In developing and carrying out their projects, researchers should include participants with germane expertise in the social and behavioral sciences, ethics, and any domains of application involved; and if their projects aim for societally relevant outcomes, they should involve relevant stakeholders throughout the research. Publications and other accounts of their research should discuss possible limitations and any downstream risks of artifacts as well as algorithms and other computing methods.

Recommendation 2. The computing research community should initiate projects that foster responsible computing research, including research that leads to societal benefits and ethical societal impact and research that helps avoid or mitigate negative outcomes and harms. Both research sponsors and research institutions should encourage and support the pursuit of such projects.

To advance such responsible, societally beneficial research projects and facilitate the multidisciplinary research called for in Recommendation 1, research sponsors and universities should support new kinds of projects and new types of partnerships with companies and philanthropies. To enable and incentivize researchers to participate in such research, (1) research sponsors should provide sufficient resources for the participation of scholars from fields other than computer science and engineering and of stakeholders, and (2) research institutions’ performance review processes and tenure and promotion committees should value both disciplinary and multidisciplinary scholarship on the ethical and societal impacts of computing research.

Recommendation 3. Universities, scientific and professional societies, and research and education sponsors should support the development of the expertise needed to integrate social and behavioral science and ethical thinking into computing research.

Academic institutions should reshape their computer science and engineering curricula and curricula in social and behavioral sciences and the humanities to better equip students to address the ethical and societal impacts of computing, and to support their graduates’ abilities to assist public and private sector institutions in making better decisions as they acquire computing technologies. Concomitantly, scientific and professional societies as well as research sponsors should provide training opportunities for computing researchers to gain proficiency in carrying out and assessing responsible computing research.

Recommendation 4. Computing research organizations–working with scientific and professional societies and research sponsors–should ensure that their computing faculty, students, and research staff have access to scholars with the expertise to advise them in examining potential ethical and societal implications of proposed and ongoing research activities, including ways to engage relevant groups of stakeholders. Computing researchers should seek out such advice.

To buttress responsible computing research efforts, research institutions, scientific societies, and research sponsors should develop ways for computing researchers to identify scholars with the ethical, societal impact, and domain expertise their projects require and provide support for such scholars to collaborate in the research. Research sponsors also should support the development and sharing of educational materials and descriptions of best practices.

Recommendation 5. Sponsors of computing research should require that ethical and societal considerations be interwoven into research proposals, evaluated in proposal review, and included in project reports.

Research proposals should describe in an integrated fashion the ethical and societal considerations associated with the proposed work and not in a separate section. Research sponsors should ensure that project review panels are provided with appropriate evaluation criteria and have the requisite expertise to evaluate these considerations, and they should require that project reports address ethical and societal issues that arise.

Recommendation 6. Scientific and professional societies and other publishers of computing research should take steps to ensure that ethical and societal considerations are appropriately addressed in publications. The computing research community should likewise take steps to ensure that these considerations are appropriately addressed in the public release of artifacts.

Conferences and journals should establish evaluation criteria and metrics to be used in assessing a paper’s treatment of ethical and societal impacts and provide guidelines for authors and reviewers. They should encourage researchers to report unanticipated ethical or societal consequences of their research and to provide guidance to future researchers interested in using the results of their research. They should also establish criteria for determining whether and how to release artifacts.

Recommendation 7. Computing researchers who are involved in the development or deployment of systems should adhere to established best practices in the computing community for system design, oversight, and monitoring.

Researchers should also be transparent about the capabilities, maturity, and limitations of any artifacts they produce and document their design assumptions.

Recommendation 8. Research sponsors, research institutions, and scientific and professional societies should encourage computing researchers to engage with the public and with the public interest and support them in doing so.

Specific opportunities for such engagement include informing the public, assisting public and private sector acquirers of computing technologies, and bringing potential adverse consequences of emerging technologies to the attention of governments and other public organizations. In addition, universities, research sponsors, and scientific societies should create opportunities for computing researchers to learn how to effectively serve in advisory capacities.

Some Personal Thoughts

This report is not about what we call Responsible Innovation (RI) or Responsible Research and Innovation (RRI) in Europe, as the idea of research aims is not addressed, only the practices. The main take-away seems to be the need for multidisciplinarity in computing research, particularly the inclusion of social and behavioural scientists and ethicists within the design and development process. The research products and processes are described as being socio-technological, and the idea that technicians can conduct responsible research without social, humanist or ethical input is criticized throughout.

There are lots of mentions of stakeholder involvement, but the idea that stakeholders can be defined and included when we are talking about computing, AI and automated decision-making would be open to criticism in a European RI discussion. The roles of funding bodies are addressed in a way similar to RI approaches, but there seems to be an underlying idea present that societies themselves will govern research, rather that research being seen as at the service of society. It is a product of society and governed by its norms, but not necessarily a tool for positive change (which is the hope that underpins the RI movement).

The report does not go as far as suggesting co-creation, but relies on language about informing the public, reporting possible problems and being transparent where possible. It also suggests limiting access for use which sems rather a problematic approach if coupled to transparency.

This is an interesting report though, a different perspective from academic or typical EU project publications. For an RI practitioner though it seems rather limited, using ideas (or at least language) from the past and not proposing any goal (or indeed questioning of) structural change. A first step into reflexivity, which is of course a positive thing in itself, but a very limited interpretation of what responsibility might be. Download it here.

———————-