Principles alone cannot guarantee ethical AI
By Brent Daniel Mittelstadt. 2019
1,403 citations recorded by the source corpus.
Artificial intelligence (AI) ethics is now a global topic of discussion in academic and policy circles. At least 84 public–private initiatives have produced statements describing high-level principles, values and other tenets to guide the ethical development, deployment and governance of AI. According to recent meta-analyses, AI ethics has seemingly converged on a set of principles that closely resemble the four classic principles of medical ethics. Despite the initial credibility granted to a principled approach to AI ethics by the connection to principles in medical ethics, there are reasons to be concerned about its future impact on AI development and governance. Significant differences exist between medicine and AI development that suggest a principled approach for the latter may not enjoy success comparable to the former. Compared to medicine, AI development lacks (1) common aims and fiduciary duties, (2) professional history and norms, (3) proven methods to translate principles into practice, and (4) robust legal and professional accountability mechanisms. These differences suggest we should not yet celebrate consensus around high-level principles that hide deep political and normative disagreement. AI ethics initiatives have seemingly converged on a set of principles that closely resemble the four classic principles of medical ethics. Despite this, Brent Mittelstadt highlights important differences between medical practice and AI development that suggest a principled approach may not work in the case of AI.
Subfields: General Health Professions, Public Health, Environmental and Occupational Health, Safety Research, Natural Language Processing, AI Ethics, Fairness & Society
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In General Health Professions
- Iason Gabriel
- Jonas Schuett
- Andreas Holzinger
- S. Lo Piano
- Aniek F. Markus
- Jan A. Kors
- Luciano Floridi
- Peter R. Rijnbeek
In Public Health, Environmental and Occupational Health
- Iason Gabriel
- S. Lo Piano
- Aniek F. Markus
- Jan A. Kors
- Luciano Floridi
- Peter R. Rijnbeek
- Thilo Hagendorff
- Luke Munn
In Safety Research
- Luke Munn
- Luciano Floridi
- Mark Coeckelbergh
- Thilo Hagendorff
- S. Lo Piano
- Hannah Rose Kirk
- Jonas Schuett
- Jakob Mokander
In Natural Language Processing
- Jey Han Lau
- Timothy Baldwin
- Kobi Leins
- Iason Gabriel
- Przemyslaw Biecek
- Wojciech Samek
- Andreas Holzinger
- S. Lo Piano
In AI Ethics, Fairness & Society
- Kobi Leins
- Iason Gabriel
- Jonas Schuett
- Ricardo A. Baeza-Yates
- Przemyslaw Biecek
- Timothy Baldwin
- Wojciech Samek
- S. Lo Piano
Citations in the corpus
- Artificial Intelligence, Values, and Alignment (2020, methodology): “19 The approach adopted by these authors has partial foundations in ‘principlism’ which is an approach commonly found in medical ethics (Beauchamp and Childress 2001; Mittelstadt 2019b)”
- Ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward (2020, background): “Seven of the most prominent value statements from the AI/ML fi elds were examined in Greene et al. (2019 Mittelstadt (2019) critically analysed the current debate and actions in the fi eld of AI ethics and noted that the dimensions addressed in AI ethics are converging towards those of medical ethics.”
- Limits and Possibilities for “Ethical AI” in Open Source: A Study of Deepfakes (2022, background): “Systems may adhere to such principles while still being patently unethical [42], and convergence on principles risks obscuring political and normative disagreements [55], or focuses “Ethical AI” scrutiny on AI design rather than the business uses it enables [30].”
- Explainable AI Methods - A Brief Overview (2020, background): “Increasing legal and data protection aspects, e.g., due to the new European General Data Protection Regulation (GDPR, in force since May 2018), complicate the use of black-box approaches, particularly in domains that affect human life, such as the medical field [56], [63], [73], [76].”
- The Ethics of AI Ethics: An Evaluation of Guidelines (2019, background): “In the light of these two approaches, the traditional type of AI ethics can be assigned to the deontological concept (Mittelstadt 2019).”
- Auditing large language models: a three-layered approach (2023, background): “Moreover, concepts like ‘fairness’ or ‘transparency’ hide deep normative disagreements [102].”
- Towards intellectual freedom in an AI Ethics Global Community (2021, background): “However, some of these initiatives in the public and private sector have been called “ethics washing” and even “ethics shopping” (Wagner [15]): ethical frameworks and initiatives are then used to avoid regulation and emphasize the role of the private sector.”
- The role of explainability in creating trustworthy artificial intelligence for health care: a comprehensive survey of the terminology, design choices, and evaluation strategies (2020, methodology): “It is difficult to ensure that these conditions hold as there are no proven methods to translate these conditions into practice [9].”
- Practical and ethical challenges of large language models in education: A systematic scoping review (2023, background): “…approach when developing and evaluating LLMs-based innovations are essential for ensuring these innovations remain ethical in practice, especially as ethical principles may not guarantee ethical AI due to their top-down manners (e.g., developed by regulatory bodies) (Mittelstadt, 2019).”
- Outsider Oversight: Designing a Third Party Audit Ecosystem for AI Governance (2022, background): “While helpful in aligning internal stakeholders, such an approach is not likely to identify or address the main concerns of external stakeholders [54, 66], particularly if that affected population or its representatives are”
- Artificial Intelligence Governance and Ethics: Global Perspectives (2019, background): “…also warranted regarding the likelihood of success for AI ethics/governance initiatives, in the form of examining the success or otherwise of previous attempts to govern new technologies, such as biotech and the Internet, or to insert ethics in other domains such as medicine (see Mittelstadt 2019).”
- A Systematic Literature Review of User Trust in AI-Enabled Systems: An HCI Perspective (2022, background): “Consequently, AI ethics guidelines can in fact be misplaced and harmful if the efforts distract the focus to operationalize ethical AI-enabled systems (Mittelstadt, 2019; Munn, 2022).”
- Give Me Convenience and Give Her Death: Who Should Decide What Uses of NLP are Appropriate, and on What Basis? (2020, background): “And even if ACL could do so, and wanted to do so, the efficacy of ethics to answer complex political and societal challenges needs to be questioned (Mittelstadt, 2019).”
- The uselessness of AI ethics (2022, background): “Mittelstadt [45, p 501] suggests that work on AI ethics has largely produced “vague, high-level principles, and value statements which promise to be action-guiding, but in practice provide few specific recommendations and fail to address fundamental normative and political tensions embedded in key…”