The Architecture of Assisted Dying: Lessons from the French Parliamentary Debates of February 2026
This article by Csongor Hetényi is a co-winner of the 2026 RJSP Prize.
The Architecture of Assisted Dying: Lessons from the French Parliamentary Debates of February 2026
In February 2026, the French Assemblée nationale cast its sounding line to plumb the uncharted depths of its law on the end of life, testing whether assisted suicide could find a place within the French medical-legal order. For years, individuals experiencing unbearable suffering had tested the Claeys-Leonetti Law of 2016,[1] which permits deep sedation and treatment refusal, only to find it inadequate. Some sought relief in Belgium,[2] while others remained in France, enduring what advocates described as a "howling existence."[3]
France does not pilot these waters alone. Eight European nations now permit some form of assisted dying.[4] However, what emerges is not a unified model but a patchwork of national compromises, each governed using rigorous safeguards. Three practices are often distinguished: active euthanasia (the physician directly ends life), passive euthanasia (withdrawing treatment), and physician-assisted suicide (providing means for the patient to act).[5] Article 2 of the French bill collapsed the first and third under "aide à mourir," a conflation that itself became a site of moral contention.
Beyond these procedural distinctions, the debate also unfolded on a deeper philosophical plane: the concept of autonomy. For proponents, autonomy grounds the right to dispose of one's own body, right up to death. For opponents, however, the individualist conception of autonomy is a myth that obscures the "unavoidably intertwined natures" of human beings.[6]
The central question is whether procedural safeguards, institutional design, or a reconceptualization of autonomy can resolve the moral tensions in assisted dying, or whether these tensions reflect irreconcilable foundational disagreements. The French debates reveal three recurring tensions, each contrasting a procedural solution with a deeper moral claim. Factual evidence supporting containment confronts the conceptual logic of expansion. The physician's conscience clause conflicts with medicine's core identity. Procedural safeguards regarding autonomy clash with a relational conception of personhood. Since these tensions stem from core disagreements, procedural improvements alone are insufficient. However, alternative institutional models, such as legal tribunals or creating a novel profession, and a reconceptualization of autonomy that includes both individual and relational dimensions, may offer solutions. This analysis unfolds in three parts, addressing safeguards, the physician's role, and the meaning of autonomy.
A. Safeguards, Normalization, and the Slippery Slope
Safeguards in assisted dying legislation serve two primary functions: guaranteeing that requests are voluntary and free from coercion, and preventing eligibility from expanding beyond the intended scope, thereby containing the feared "slippery slope."
In the French bill, proponents point to the five cumulative eligibility criteria in Article L. 1111-12-2 as evidence of thorough calibration.[7] The article requires the person 18 years old (1°), of French nationality or stable residence (2°), suffering from a serious and incurable condition with an advanced or terminal vital prognosis (3°), presenting constant physical or psychological suffering that is refractory or unbearable, with the explicit limitation that "une souffrance psychologique seule ne peut en aucun cas permettre de bénéficier de l'aide à mourir" (4°), and capable of manifesting their will freely and in an informed manner (5°). The procedure requires a written request,[8] a collegial process with at least one independent physician,[9] a 15-day decision window, [10] a 2-day reflection period,[11] and post hoc review by a commission.[12] As Agnès Firmin Le Bodo said, this is "not a wide-open door" but "a marked path" where the patient is "accompanied and controlled."[13]
Opponents challenge these safeguards. The British Medical Association warns that safeguards are "simply statements of what should happen in an ideal world," failing to reflect real-world clinical and familial dynamics.[14] Richards adds a theoretical critique: safeguards merely describe who currently qualifies; as the conception of dignity shifts, the boundaries will shift too.[15] Echoing Richards, Christophe Bentz called the criteria "safeguards of the moment" that "will be removed in the months and years to come."[16]
Empirical evidence from the Netherlands and Belgium offers an opposing narrative. Dutch data show that life-terminating acts free of explicit patient request decreased from 0.8% to 0.4% of deaths between 1990 and 2005; the authors concluded that "the legalization of euthanasia in the Netherlands did not result in a slippery slope."[17] The review found "no evidence for a higher frequency of euthanasia" among fragile populations.[18] Belgian data confirm these outcomes: Jacques Wels and Natasia Hamarat found that cases involving psychiatric and cognitive conditions remained stable.[19] As Luc Deliens summarizes, research from Belgium consistently shows that vulnerable people do not have easier access, and the highest request rates are found among those with higher educational degrees, not among the poor or less educated (Deliens, 2025).[20]
A slippery slope proponent might object that Dutch and Belgian laws differ from the French bill. The Netherlands has no temporal eligibility criterion; [21] Belgium uses "short term" only as a procedural trigger, not as an eligibility condition.[22] Both permit psychological suffering alone,[23] while the French bill explicitly excludes it.[24] However, this objection overlooks further structural similarities: all three require a serious, incurable condition;[25] unbearable suffering;[26] a voluntary, well-considered request;[27] and independent medical consultation.[28] The Dutch and Belgian data are therefore the best available evidence on whether well-designed safeguards can contain expansion, though cautious inference is warranted given differences in minor eligibility and psychological suffering.
A second objection is that the absence of detected expansion does not prove expansion cannot occur; the slippery slope may be gradual and qualitative, not captured by aggregate statistics. This is a valid limitation. The empirical evidence does not refute Richards' conceptual claim, it only shows that after two decades, the feared expansion to at-risk populations has not occurred. Whether this refutes the claim depends on whether one believes the logic of expansion is inevitable or merely possible.
Thus, the debate reveals an unresolved tension: empirical evidence suggests well-designed safeguards can contain expansion, yet conceptual critiques warn the law's internal logic may pressure expansion that safeguards alone cannot indefinitely restrain. Whether the French model will replicate the stable outcomes in the Netherlands and Belgium remains an open question. Even if safeguards contain expansion, a deeper problem persists: the role of physicians and whether doctors should be agents of death. The analysis now turns to this question.
B. The Role of the Physician
Opponents of the French bill argued that authorizing assisted dying conflicts with medicine's core identity. Matthias Renault invoked the Hippocratic oath: "I will never deliberately cause death."[29] Claire Marais-Beuil cited a physician's consultation in which "66 percent of the responses … were negative" when asked whether doctors should administer lethal substances.[30] The BMA warned that licensing doctors to kill "crosses a Rubicon in medicine."[31] These objections share a belief that authorizing assisted dying would irrevocably alter the doctor's role from healer to killer.
Proponents countered that the physician's role had evolved before. The Royal Society of Canada Expert Panel noted that "the practice of medicine has evolved … decisions to terminate treatment are routinely taken" and that "it is uncontroversial … that the healthcare professional's role is not limited to providing therapy."[32] Agnès Firmin Le Bodo defended the bill's conscience clause (Art. L. 1111-12-12), which allows any health professional to refuse to participate and requires referral to willing colleagues, arguing that the distinction between self-administration and third-party administration "is … an ethical difference, thanks to which health care professionals can have confidence in the system."[33]
Three answers emerge to the question of whether medicine is the appropriate domain for assisted dying. Fiona Randall and Robin Downie argue that assisted dying should be removed from healthcare entirely. The General Medical Council defines death as a "serious adverse outcome" of treatment.[34] From this, they conclude that " if a doctor were to prescribe a lethal medication with the aim of assisting a patient to kill themself then the doctor would be aiming at an ‘adverse outcome’. But it makes no sense to say that an adverse outcome could be an aim of treatment."[35] They propose that a legal tribunal should make the final decision, and non-medical technicians should administer the lethal substance.[36] Doctors would be involved only in confirming diagnosis and prognosis.[37]
Barker and colleagues offer a second answer: a novel profession should be created. They argue that requiring doctors to act as agents of death "disrupts the defining characteristic of being a doctor."[38] Instead, "society would be more appropriately served by considering the role of an [assisted dying practitioner] as a novel profession for which the goals, competencies, research base and regulation can be established independently."[39] This profession would be philosophically and legally singular, distinct from medicine, with its own training and ethical framework.
Greta Seveso and colleagues offer a third answer, challenging the premise that physicians must be removed. Drawing on evidence from Belgium and the Netherlands, they argue that "it is crucial to reframe the concept of care to include alleviating suffering, even when it involves assisting in a patient's choice to end their life."[40] A broader understanding of care, such as supporting patients in their autonomous decisions about death, can reduce the psychological burden on professionals. Moreover, opposition often stems from inadequate preparation rather than inherent incompatibility. "Without adequate preparation, physicians and other medical staff may struggle with ethical uncertainties, fear of making procedural errors, or difficulties in communicating effectively."[41] Countries with well-established assisted dying laws have implemented consultation services, debriefing sessions, and mentoring programs that enable physicians to participate without compromising their professional integrity.[42]
A French proponent might object that the bill limits physicians to validating diagnosis and prognosis, not administering the lethal substance (except when the patient is physically unable). This is closer to the Randall and Downie model than full physician administration. However, even diagnosis and prognosis are not value-neutral when the patient knows the diagnosing physician may later facilitate death. The therapeutic relationship is altered by the possibility of the physician acting as gatekeeper to lethal drugs. Moreover, the conscience clause does not shield the profession from the cultural consequence: the public will still see doctors as those who provide death, regardless of individual opt-outs.
The French bill's model, positioning physicians as gatekeepers rather than administrators, sits uneasily between these approaches. It neither fully removes physicians nor adequately supports them. Its success depends on adopting the training and support mechanisms identified by Seveso et al. as essential. Without them, the alternatives proposed by Randall and Downie or Barker et al. may be better.
C. What is Autonomy?
Richards contends that "a society enamored by individualism will uphold an individualistic autonomy as obvious and will be blind to even seeing its intrinsic inadequacies,"[43] while proponents of relational autonomy argue for a more socially embedded conception of the self. This tension between individualist and relational conceptions of autonomy is not unique to France; it recurs across Western jurisdictions, where the liberal individualist tradition predominates in law and ethics. Although particular legal rules differ among countries, procedural safeguards alone cannot resolve this tension. Addressing it requires a fundamental reconceptualization of autonomy.
The classic individualist conception defines autonomy as "the ability to make individual, fully-informed, and independent decisions"[44] From this perspective, the moral legitimacy of assisted dying follows directly: a self-owning, rational, and informed individual has a liberty right to choose the timing and manner of their death. In the French debates, this system found forceful expression. René Pilato grounded the right in self-ownership: "To whom does a human being's body belong? … It belongs to the consciousness that animates it. Everyone must therefore be able to dispose of it until the last second."[45] Rapporteur Élise Leboucher called aid in dying "a new liberty … a right to self-determination,"[46] while Ségolène Amiot linked the struggle to prior social movements against "the domination of society that opposes the liberty to dispose of one's body."[47]
Opponents challenged this individualist conception. Thibault Bazin argued that "the balance that will gather a majority of support is not necessarily an ethical balance, which would be achieved if the ethics of autonomy did not neglect the ethics of vulnerability, equally respectable."[48] Annie Vidal articulated a collective vision of "accompaniment" and "a promise of non-abandonment made by families … by caregivers … by the fraternal Republic."[49] These critiques point toward relational autonomy, which views the person as socially embedded and interdependent.[50] Studies show that focusing solely on individual autonomy does not reflect patients' actual end-of-life experiences[51] and that individual preferences cannot be separated from the environments in which they are embedded.[52]
A particularly incisive critique comes from Anita Ho, who argues that the individualist model "neglects the social structure within which health-care decisions are made."[53] The deeper flaw, for Ho, is that individualistic autonomy "focuses on whether an individual is making a rational decision among the available options without considering if some desired alternatives have already been restricted by the social structure."[54] In fact, court cases show that many who seek assisted death would want to live if social services and opportunity-enhancing arrangements were available.[55] Thus, the choice to die is not truly autonomous if it proceeds from a lack of institutional support rather than from an uncoerced preference. Better social services would lead many who currently seek death to choose life.[56]
This institutional critique applies directly to France. As Sandrine Dogor-Such noted, the Leonetti laws are not fully applied, and "nearly 50% of people who need palliative care do not have access to it."[57] Without guaranteed access to palliative care, the choice to seek assisted dying may be driven by institutional failure rather than autonomous preference.
Beyond material resources, relational autonomy also addresses social-psychological perceptions. Charles de Courson warned that fragile persons might choose death not freely but because they "perceive themselves as a burden on their family, through solitude, through weariness, through despair."[58] Vincent Trébuchet added: "by judging implicitly that certain deaths are undignified, you make dignity a prize to be won … Nothing wounds equality more deeply."[59] Thus, the desire to die is not purely a function of impairment or suffering, but also of internalized social messages about worth and cost. A relational approach requires challenging the attitudes that make people feel like burdens rather than merely providing resources.
Crucially, relational autonomy does not reject autonomy but reconceptualizes it. Others' influence does not necessarily impede autonomy but can enhance it.[60] Healthcare professionals and family members can improve decision-making by presenting alternatives,[61] providing emotional support,[62] removing social barriers,[63] or bridging gaps between the patient and social environment.[64] Thus, the default should be to see relations as positive forces rather than threats, focusing on properly articulating each person's role rather than "freeing" the patient from their social environment.[65]
Turning to legal implementation, current legal standards align with an individualistic view, and Western legal systems are insufficiently sensitive to collective approaches.[66] Practical suggestions include familial advance directives, documents signed by the patient together with the family, and community-based informed consent documents that consider relatives' influence. These are based on moderate familialism, where the family has "default but not the absolute authority in the decision-making process."[67]
Finally, some critiques must be anticipated. The main concern is protecting patients against unwarranted family interventions.[68] Moreover, many authors present relational autonomy as the mere "mirror" of individualist autonomy, making dialogue unfruitful.[69] Yannick Monnet captured this nuance, observing that questions on autonomy resist simplistic categorization into opposing camps.[70] How to resolve conflicts when individual preference and relational influence diverge remains an open question. The way forward is further development of autonomy that considers both dimensions: no relationality without individuals; no individuality without relations.
Conclusion
In February 2026, the French Assemblée nationale sounded the depths of its law, but the resulting chart was marked by blank spaces and uncertain contours. The debates demonstrate that assisted dying legislation, like any legal framework, has inherent boundaries. Three fundamental tensions emerged. First, empirical evidence from the Netherlands and Belgium indicates that well-designed safeguards can contain the slippery slope for two decades,[71] yet conceptual critiques highlight the law's internal pressures toward expansion. Second, authorizing physicians to end life challenges medicine's core identity as healer, while removing physicians from the process, prompts questions about appropriate agents for assisted dying. Third, the individualist conception of autonomy, which supports a right to die, conflicts with a relational conception of personhood that stresses a duty of care.
The results indicate that procedural safeguards alone cannot fully resolve the ethical conflicts inherent in assisted dying. Legal provisions cannot address issues that are fundamentally philosophical. For legislators, this underscores that legal design must be complemented by solid palliative care, transparent monitoring, and ongoing ethical deliberation. In moral philosophy, the debates highlight the need for a dual-dimensional conception of autonomy that accepts both individual self-determination and social embeddedness. While the French debates did not resolve these questions, they offered an essential springboard for future exploration.
[1] Law No 2016-87 of 2 February 2016, creating new rights in favor of patients and persons at the end of life.
[2] Commission Fédérale de Contrôle et d'Évaluation de l'Euthanasie (Distelmans W, Hamarat N, Herremans J and Proot L) Onzième rapport aux Chambres législatives (années 2022-2023) (Brussels 2024) 34.
[3] Olivier Falomi, XVIIth legislature, Ordinary session of 2025-2026, Second sitting of Monday 16 February 2026 (2026).
[4] Blaakman S, Euthanasia Legislation in the EU (European Parliamentary Research Service, Members' Research Service, PE 775.914, September 2025) 1, 1.
[5] Fontalis A, Prousali E and Kulkarni K, 'Euthanasia and Assisted Dying: What Is the Current Position and What Are the Key Arguments Informing the Debate?' (2018) 111(11) J R Soc Med 407, 407.
[6] Richards S, 'The Morality of Assisted Dying' (2025) 50(4) Journal of Medicine and Philosophy 262, 276.
[7] Bill No 2401, relating to the right to assisted dying (28 January 2026), Art L 1111-12-2 (1°-5°).
[8] ibid, Art L 1111-12-3, I.
[9] ibid, Art L 1111-12-4, II.
[10] ibid, Art L 1111-12-4, III.
[11] ibid, Art L 1111-12-4, IV.
[12] ibid, Art L 1111-12-13.
[13] Agnès Firmin Le Bodo (n 3).
[14] British Medical Association, Key Arguments Used in the Debate on Physician-Assisted Dying (BMA 2021) 1, 1.
[15] Richards (n 6) 277-78.
[16] Christophe Bentz, XVIIth Legislature, Ordinary Session of 2025-2026, First Sitting, Thursday 19 February 2026 (2026).
[17] Rietjens JAC and others, 'Two Decades of Research on Euthanasia from the Netherlands: What Have We Learnt and What Questions Remain?' (2009) 6(3) Journal of Bioethical Inquiry 3, 6.
[18] ibid 6.
[19] Wels J and Hamarat N, 'Incidence and Prevalence of Reported Euthanasia Cases in Belgium, 2002 to 2023' (2025) 8(4) JAMA Network Open 1, 1.
[20] Deliens L, 'Assisted Dying and the Slippery Slope Argument: No Empirical Evidence' (2025) 8(4) JAMA Network Open 1, 1-2.
[21] Government of the Netherlands, Wet toetsing levensbeëindiging op verzoek en hulp bij zelfdoding (Termination of Life on Request and Assisted Suicide (Review Procedures) Act) (2002), Art 2.1.
[22] Belgian Parliament, Loi relative à l'euthanasie (The Belgian Act on Euthanasia) (28 May 2002), Art 3.3.
[23] ibid, Art 3.1; Netherlands (n 21), Art 2.1.
[24] Bill No 2401 (n 7), Art L 1111-12-2, 4°.
[25] Netherlands (n 21), Art 2.1; Belgian Parliament (n 22), Art 3.1; Bill No 2401 (n 7), Art L 1111-12-2, 3°.
[26] Netherlands (n 21), Art 2.1; Belgian Parliament (n 22), Art 3.1; Bill No 2401 (n 7), Art L 1111-12-2, 4°.
[27] Netherlands (n 21), Art 2.1; Belgian Parliament (n 22), Art 3.1; Bill No 2401 (n 7), Art L 1111-12-2, 5°.
[28] Netherlands (n 21), Art 2.1.e; Belgian Parliament (n 22), Arts 3.2, 3.3; Bill No 2401 (n 7), Art L 1111-12-4, II.
[29] Matthias Renault (n 16).
[30] Claire Marais-Beuil (n 16).
[31] British Medical Association (n 14) 1.
[32] Schüklenk U and others, 'End-of-Life Decision-Making in Canada: The Report by the Royal Society of Canada Expert Panel on End-of-Life Decision-Making' (2011) 25(S1) Bioethics 1, 46.
[33] Agnès Firmin Le Bodo (n 3).
[34] Randall F and Downie R, 'Assisted Suicide and Voluntary Euthanasia: Role Contradictions for Physicians' (2010) 10(4) Clinical Medicine 323, 324, citing General Medical Council, Consent: Patients and Doctors Making Decisions Together (GMC 2008).
[35] ibid.
[36] ibid 326.
[37] ibid.
[38] Barker S, Fritz Z and Ruck Keene A, 'Why Administration of Lethal Drugs Should Not Be the Role of the Doctor' (2026) 52 Journal of Medical Ethics 1, 1.
[39] ibid.
[40] Seveso G and others, 'Medically Assisted Suicide in Italy after Regional Law No 16/2025: Perspectives, Challenges, and the Need for Psychological Support and Specialized Training for Healthcare Professionals' (2025) 0(0) Medicine, Science and the Law 1, 5.
[41] ibid 6.
[42] ibid.
[43] Richards (n 6) 279.
[44] Głos A, 'Solidarity in Healthcare: The Challenge of Dementia' (2016) 49 Diametros 1, 3.
[45] René Pilato (n 16).
[46] Élise Leboucher (n 3).
[47] Ségolène Amiot (n 16).
[48] Thibault Bazin (n 16).
[49] Annie Vidal (n 3).
[50] Gómez-Vírseda C, De Maeseneer Y and Gastmans C, 'Relational Autonomy: What Does It Mean and How Is It Used in End-of-Life Care? A Systematic Review of Argument-Based Ethics Literature' (2019) 20 BMC Medical Ethics 1, 9.
[51] Virdun C and others, 'Dying in the Hospital Setting: A Systematic Review of Quantitative Studies Identifying the Elements of End-of-Life Care that Patients and Their Families Rank as Being Most Important' (2015) 29(9) Palliative Medicine 774, 777.
[52] Ikonomidis S and Singer PA, 'Autonomy, Liberalism and Advance Care Planning' (1999) 25(6) Journal of Medical Ethics 522, 524; Asagumo A, 'Relational Autonomy, the Right to Reject Treatment, and Advance Directives in Japan' (2022) 14 Asian Bioethics Review 57, 65.
[53] Ho A, 'The Individualist Model of Autonomy and the Challenge of Disability' (2008) 5(2-3) Journal of Bioethical Inquiry 193, doi:10.1007/s11673-007-9075-0.
[54] ibid 203.
[55] Ho A, 'Disability in the Bioethics Curriculum' (2007) 30(4) Teaching Philosophy 403, 412.
[56] Ho (n 53) 202.
[57] Sandrine Dogor-Such (n 16).
[58] Charles de Courson (n 3).
[59] Vincent Trébuchet (n 3).
[60] Gastmans C and others, 'From Birth to Death? A Personalist Approach to End-of-Life Care of Severely Ill Newborns' (2013) 19(1) Christian Bioethics: Non-Ecumenical Studies in Medical Morality 7, 12.
[61] Walter JK and Ross LF, 'Relational Autonomy: Moving Beyond the Limits of Isolated Individualism' (2014) 133(Suppl 1) Pediatrics S16, S22.
[62] Grignoli N, Di Bernardo V and Malacrida R, 'New Perspectives on Substituted Relational Autonomy for Shared Decision-Making in Critical Care' (2018) 22(1) Critical Care 1, 3.
[63] Ho (n 53) 204.
[64] Schicktanz S and Schweda M, 'The Diversity of Responsibility: The Value of Explication and Pluralization' (2012) 3(3) Medicine Studies 131, 131.
[65] Gómez-Vírseda, De Maeseneer and Gastmans (n 50) 9.
[66] Gilbar R and Miola J, 'One Size Fits All? On Patient Autonomy, Medical Decision-Making, and the Impact of Culture' (2015) 23(3) Medical Law Review 1, 4.
[67] Gómez-Vírseda, De Maeseneer and Gastmans (n 50) 10.
[68] Grignoli, Di Bernardo and Malacrida (n 62) 2.
[69] Gómez-Vírseda, De Maeseneer and Gastmans (n 50) 11; Ikonomidis and Singer (n 52) 525.
[70] Yannick Monnet (n 3).
[71] Rietjens and others (n 17) 3, 6.
Commentaires