Chastening and enthusiasm about genome editing

A writer in Nature says that China sent a “strong signal” by punishing He Jiankui and two colleagues with fines, jail times, and bans against working again in human reproductive technology or applying for research funding.  (They lost their jobs as well and may not be able to do research work, presumably in any field, in a Chinese institution again.)  It is encouraging, this writer says, that China took this action demonstrating a commitment to human research ethics.  He and other researchers doing gene-editing work that is not ethically objectionable worry that there may be collateral damage, so to speak, against ANY gene-editing research in China.

Another writer in Nature says cites progress under “appropriate caution” for using gene editing techniques for so-called “somatic” gene editing; that is, editing disease genes in an existing person with that disease, to treat it.  This is, in essence, a form of gene therapy and is ethically permissible under proper research ethical guidelines.  Some clinical trials in progress involve injecting the gene-editing apparatus into a person, while most such trials remove the person’s blood cells, edit them in the laboratory, then re-introduce them into the bloodstream, after which the edited cells are left to mature normally.  The latter approach is particularly attractive to treat genetic blood diseases such as sickle cell anemia.

Both perspectives seem correct, as far as they go—never mind whether Dr. He’s jail sentence fits the crime, as Joy Riley asked on this blog last weekend.  Never mind also whether Dr. He’s research should be published; as Mark McQuain commented, it’s a bit incongruous to want to assess the technical merits of work that should not have been done in the first place.  He linked an opinion in Technology Review that argued, briefly, that because the ethics of editing genes in human embryos is under societal debate, people trying to decide on the ethical merits should be able to assess for themselves whether Dr. He succeeded, technically at what he set out to do.  (The consensus to date seems to be, no, he did not.)   But the role of technical success in assessing the ethical merits of a medical intervention—or, better, an intervention made in the name of medicine—depends on the degree to which the ethical judgment is a matter of making a reasonably reliable of risk and benefit, and the degree to which risk-benefit is a criterion for judging the ethical merits.  And therein, as they say, lies the rub—which I hope to revisit in coming posts.

2020, or 20/20?

Near the end of 2018, He Jiankui was on the world’s stage announcing that he had edited the genome of twin girls, in the hope of making them resistant to HIV. On Tuesday, December 31, 2019, the Wall Street Journal (WSJ) printed a report that Dr. He and two others have been convicted of “illegally practicing medicine related to carrying out human-embryo gene-editing intended for reproduction.” (online version here).

A court in Shenzhen concluded that the defendants had acted for “fame and profit,” when they “deliberately violated the relevant national regulations, and crossed the bottom lines of scientific and medical ethics.” For the crime committed, He received the most severe sentence. In addition to the three-year prison sentence, He is banned for life from “working in the field of reproductive life sciences and from applying for related research grants, “ according to the WSJ.

The Xinhua News Agency also noted that a third genome-edited baby had been born, and that this child, along with the previously born twins, “would be monitored by government health departments.” The WSJ did not state for how long the monitoring would continue. Not only were the children experimental subjects as embryos, but they continue to be subjects as well. Further, these genome effects will affect their progeny, potentially into perpetuity. Additionally, the Smithsonian Magazine reports that in the summer of 2019, He met with “investors to discuss a potential commercial genetic modification clinic in Hainan, which aims to become a ‘world-class medical tourism hub’.”  One might reasonably call this “a crime against humanity,” even if it does not include genocide of humans already born. (For further reading, see David Luban, “A Theory of Crimes Against Humanity”)

In the print edition of the WSJ, alongside the article on He is an article about Pastor Wang Yi of the Early Rain Covenant Church. Pastor Wang was sentenced on 30 December to nine years in prison. His crime was “incitement of subversion of state power and illegal business operations” (online article here).

Consider that a pastor receives a nine-year sentence for an offense against the State; and a scientist, a sentence of three years for a crime against multiple generations, and indeed, humanity. In the year 2020, we could use a check of our understanding of what is important in the life of the world. Would that our vision were 20/20 also.

Experimental Subjects for Life?

More than a year after the birth announcement of genome-edited babies in China, we are only slightly more informed of He Jiankui’s experimentation, the results of which are named “Lulu” and “Nana.” Although apparently approached, neither Nature nor the Journal of the American Medical Association (JAMA) chose to publish He’s work. Antonio Regalado reported on an unpublished manuscript in “China’s CRISPR babies: Read exclusive excerpts from the unseen original research” in Technology Review on 3 December 2019. The Technology Review article includes not only excerpts of the manuscript from He, but also reactions from Stanford law professor Hank Greely; University of California—Berkeley’s gene-editing scientist Fyodor Urnov; the scientific director of Eugin assisted reproduction clinics, Rita Vassena; and reproductive endocrinologist Jeanne O’Brien, from Shady Grove Fertility.

Regalado summarizes some of the problems with Chinese experiment as follows:

 . . . key claims that He and his team made are not supported by the data; the      babies’ parents may have been under pressure to agree to join the experiment; the supposed medical benefits are dubious at best; and the researchers moved forward with creating living human beings before they fully understood the effects of the edits they had made.

Greely points out the lack of “independent evidence” of the claims made in the paper. Urnov labels the paper’s claim of reproducing the usual CCR5 variant “a deliberate falsehood,” and calls the statement about the possibility of millions being helped through embryo editing “equal parts delusional and outrageous.” O’Brien’s concerns include the possibility of coercion of the couples involved, and, noting the social stigma of HIV-positivity in China, she poses the question of whether this was a genetic fix for a social problem. Certainly, the Chinese experiment raises many questions, including how a culture views children. Are children gifts to be received or projects to be completed? Is it appropriate to subject children to experimental research because we can? One of the quotes from the paper reads, “we have made a follow-on plan to monitor the health of the twins for 18 years and hope to then reconsent for continued monitoring through adulthood.”

We would be remiss if we thought that China alone plans to remake humanity. Vassena is quoted regarding He’s study:

Unfortunately, it reads more like an experiment in search of a purpose, an    attempt to find a defensible reason to use CRISPR/Cas9 technology in human embryos at all costs, rather than a conscientious, carefully thought through, stepwise approach to editing the human genome for generations to come.  As the current scientific consensus indicates, the use of CRISPR/Cas9 in human embryos destined to give rise to a pregnancy is, at this stage, unjustified and unnecessary, and should not be pursued.

Vassena, who directs a fertility enterprise, it should be noted, appears comfortable with impacting the human genome for generations to come:  It just needs to be a “reflective” and “mindful” approach. That is chilling. Would she, or the study’s authors, or Greely, or Urnov, or O’Brien sign up to be a science experiment for the rest of their lives? I would not consent—not for myself nor for my children—no matter how “reflective” or “mindful” the researcher happened to be.

Finally, “Lulu” and “Nana” should be known as more than the results of someone’s laboratory experiment. They are human beings, not laboratory rats or cells under a microscope to be studied at the will and convenience of the experimenters.

“Why did you make me this way?!”

Recently, Jon Holmlund brought us up to date on an effort in Russia to proceed with CRISPR gene editing aimed at eliminating deafness. Coincidently, a recent MedPage article was posted regarding the ethics of using pre-implantation genetic diagnosis and IVF to purposefully select FOR an embryo with genetic deafness for a couple, both of whom were deaf. Both links discuss some of the ethical problems with using medical reproductive and genetic technology for these purposes. While we presently lack the ability to use polygene scoring to accurately “produce” the babies we want, I want to use the remainder of today’s blog to consider what obligations, if any, a genetic engineer (or parents that use their skills) may have toward future children designed using these growing array of genetic technologies.

Deafness seems to me to be rather curious in that it is considered either a serious disability or a desirable trait, depending upon your cultural worldview. No one in any culture would purposefully select for cystic fibrosis or Tay-Sachs diseases for their child. In fact, most want to use medical reproductive and genetic technologies to eliminate these diseases. On the other end of the genetic trait spectrum, some parents want to use these same technologies to purposefully select for more trivial traits for their children – hair and eye color, for instance. Given the triviality of these traits, I hear no one mounting an effort to genetically eliminate any particular hair and eye color. Perhaps I am living a sheltered life?

Nonetheless, with regard to deafness, prior to the promises of our new reproductive technologies, if you were born deaf and did not like it, you could only shake your fist at God or Nature. Now (or very soon), you can shake your fist directly at another human, such as your regional genetic engineer (or your parents who purposefully used her technological skills) and demand a direct answer as to why they purposefully made (or did not make) you deaf. Maybe this angst will be more widespread for the many more trivial traits such as eye or hair color rather than something more significant like deafness?

Building a child is about to become much harder for parents as they become directly responsible for both interior (genetic illnesses) and exterior (hair and eye color) design issues.

“Why DID you make me this way?!”

“Why did YOU make me this way?!”

I wonder how many genetic designers (or the parents that will ultimately bear the direct responsibility for having used the technology) really want that type of responsibility?

Is there already fine print in PGD-IVF contracts holding the doctors/scientists/geneticists harmless for the choices the parents make?

I can’t wait for the late night TV commercials: “Were you born with brown eyes and feel emotionally scarred because you have always wanted blue? Call our law offices as you may be entitled to financial compensation …”

The importance of premises

In an interesting article in the Hastings Center Bioethics Forum, titled “Hannah Arendt in St. Peter’s Square,” Joseph Fins and Jenny Reardon write about the importance of deep ethical reflection in dealing with the ethical challenges of biomedical research. They point out that when ethics becomes a matter of simply following a set of rules we can end up in the wrong place. Even such fundamentally good concepts as informed consent and the need to have research proposals reviewed to be sure that they are ethically sound can lead to a mindset of regulatory compliance, essentially following the letter of the law, while leading to poor conclusions about what we ought to do. In the end they suggest that in order to facilitate deeper ethical thinking regarding new areas of biomedical research we need more interdisciplinary conversation between the sciences and engineering on one hand and the humanities and social sciences on the other. I think this is quite true and is a strong argument for a liberal education in its classic sense.

However, I find it particularly interesting how the thinking of Hannah Arendt enters into their discussion. Arendt was a German Jew who fled from Europe to the US in the Nazi era. She wrote about the kind of thinking that allowed the totalitarian regimes of Hitler and Stalin to gain control. Fins and Reardon focus on her idea that logical thinking can lead from a seemingly self-evident statement to a replacement of common sense with thinking that leads in a direction that is very wrong. They see a culture in medicine and science that considers ethics as a matter of regulatory compliance rather than deep reflection an example of this.

What I find most interesting in Arendt’s thinking is the idea that logic will lead to faulty conclusions if the premise is not true. The problem that she saw in the thinking leading to totalitarian regimes was not that the thinking was illogical. The problem was that the seemingly self-evident statements which were used as the premises were false. When we apply that to ethics it means that we will only reach sound ethical conclusions when we begin with moral premises that are true. A liberal education with interplay between the humanities and the sciences is one way to seek true premises for our ethical thinking in the wisdom that can be found in the interplay of academic disciplines. Another is to recognize that the existence of common sense morality suggests a source of moral wisdom that is beyond human wisdom. Christian ethics finds its premises in that higher source of moral wisdom. A Christian liberal education integrates them both.

Stem Cell Rx No Longer For Sale on Google

Perhaps once a week, I will be asked by a patient about the potential benefits of stem cells for reversing the normal affects of age, particularly with respect to arthritis of the knee joints, hip joints or the degenerative discs in the lumbar spine. I believe one of the reasons for this interest has come from increasing advertisements by various clinics in my region of East Tennessee claiming stem cells are the answer for these problems. My region is not unique. A simple Google search on “stem cells for knee pain” yields ads for clinics offering such treatment.

Stem cells are cells that have potential to become any type of cell in the human body such as a new blood cell, nerve cell or bone/cartilage cell. Scientists are rapidly learning how to find or create stem cells, as well as how to safely use them to replace old or missing cells, thus restoring function in worn out, damaged or diseased areas of the body. In fact, stem cells are presently used to replace the bone marrow for some individuals with certain cancers and disorders of the blood and immune system, and in many of these cases, the results are lifesaving.

The problem is that stem cell treatment remains yet unproven in all other medical conditions, including the age-related arthritis conditions which I treat. This lack of efficacy has not stopped clinics from offering and patients from receiving stem cell injections with the hope of achieving improved function or cure. I am willing to grant that many offer these treatments with the sincere hope and belief that they are acting in their patient’s best interest, though I suspect not all have the patient’s best interest in mind. Unfortunately, there have been severe adverse events. Examples include blindness following an injection of stem cells into the eye, and loss of function with development of a spinal cord tumor following stem cell injection into the spine.

The FDA is trying to educate the public and prevent stem cells from being offered for unproven treatments. The FDA has the authority in the US to stop these unproven treatments and take punitive action if needed. This is not to suggest that the FDA is in the business of preventing legitimate investigation into the potential benefits of stem cells, such as this Mayo Clinic Phase 1 study looking at the risks of injecting stem cells in to the cerebrospinal fluid of patients following a spinal cord injury to see if this particular stem cell technique causes harm (with future studies needed to determine benefit).

The FDA is recently getting some help from Google. On September 6th, Google announced it would stop accepting ads for unproven medical treatments, including stem cell therapies. It is early in the effort and the initial link above still has four ads for non-bone marrow stem cell treatments returned with the Google search. Maybe by the time you read this blog entry, the stem cell ads for unproven treatments will be gone.

I am hopeful that stem cells will eventually provide patients with safe therapies that repair injury and return patients to normal health. Offering that promise without the studies that prove such benefit is unethical and potentially harmful. It is good to see Google favoring human welfare over financial profit.

Promoting vaccination with a not-too-heavy hand

This week’s Nature has a worthwhile read, “Mandate Vaccination with Care.”    The recent rise in the number of cases of measles is well-documented in the general press, and there is a strong argument that it is a social good that sufficient numbers of children be vaccinated for a range of infectious diseases.  Your correspondent considers it unfortunate, to put it mildly, that there is a persistent belief that vaccines for the standard childhood diseases are harmful.  Although some cases of vaccine harm occur, they are rare—rarer than many in the general public believe—and the cost of under-vaccination is great.  I, for one, never want to see an infant with pertussis (whooping cough) again, and, although I recall having had measles and chicken pox when I was a kid, it’s best to prevent them.  Some can even be eradicated (see: smallpox—which we should fervently hope is never purposely re-introduced, now that we don’t routinely vaccinate for it). 

In brief, the authors in this case argue for promoting vaccination in the public with such steps as ensuring supply and access, providing information and allowing public forums, monitoring safety carefully, and tracking vaccination rates.  They argue, reasonably, that mandatory vaccination that carries the wrong kind of penalties—such as, fines or even jail sentences imposed in some countries—for non-compliance actually can harm poorer, medically underserved people, and as such be counterproductive and, frankly, unfair.  They comment that harsh mandates can unnecessarily prompt a backlash, with increased resistance.

They say, further, that if mandates are deemed “politically appropriate,” then the procedures should be just, with constraints on choice as limited as possible; any penalties must be proportionate; those who do suffer complications should be adequately compensated.   They speak favorably of creating administrative hurdles to getting exemptions from mandates.  They also argue against governments mandating only some vaccines while excluding others.  They claim that making some vaccines only “recommended” can limit the uptake of all. 

This last point may be the most questionable of all in this article.  It is easier to justify mandating vaccination for highly contagious diseases that can have devastating effect (e.g. measles, rubella, diptheria, and others), than, for example, vaccination for human papilloma virus (HPV), infection with which predisposes to certain kinds of cancer but transmission of which is through sexual activity.  In this last case, the argument for a mandate is substantially weaker; vaccination at a fairly young age might be wise, but one might still reasonably accept, for oneself or one’s child, the less certain and more remote risks of the consequences of infection, and therefore reasonably object to mandated vaccination.

Again, a worthwhile read. >

Good from Evil

I was given an article by a student of mine following his one month elective rotation with me in which we spent some clinical time discussing bioethical issues. The May 2019 web article by Sharon Begley from Statnews.com had to do with an interesting medical dilemma first presented in 2016 by Dr. Susan Mackinnon from Washington University in St. Louis. I have briefly summarized Begley’s article in the first part of today’s blog and extended her point at the end.

Dr. Mackinnon had a patient who was having severe leg pain following multiple knee surgeries. Dr Mackinnon was providing the final surgical attempt to isolate the nerve presumably being compressed by scar tissue in hopes of surgically decompressing that nerve to permanently relieve the patient’s severe pain. If the surgery was not successful, the only other option at that point was to amputate the leg. During the surgery, she used an old anatomy book called The Pernkopf Topographic Anatomy of Man, which unambiguously has the best illustrations of nerves around the knee, and successfully located and decompressed the nerve in question and successfully avoided an amputation.

So, what was the dilemma?

As Begley points out in her article, it came to light in the mid-1980s that the illustrations used in the Pernkopf atlas were based in part on the bodies of people executed by the Nazis in the late 1930s. The moral dilemma for Dr. Mackinnon was therefore:

“…even now, the Pernkopf illustrations are unsurpassed in their accuracy and detail, especially their depiction of peripheral nerves…and although a few journal papers may have an equally good, single illustrations, finding the right paper takes time that Mackinnon did not have as she stood over her patient.”

Dr. Mackinnon had been given the Pernkopf atlas as a graduation gift in 1982 but the Nazi history behind the atlas was not known until the mid-to-late 1980s, the full history of which only became known to her after the surgery. Should she continue to use an atlas that contains illustrations of the bodies of people executed by the Nazis? If used, is there a duty to inform a current patient about the nature of the atlas? Can sufficient good be derived from the atlas given the unspeakable evil required to create it to permit its ongoing use?

She posed her dilemma to Rabbi Joseph Polak, the Chief Justice of the Rabbinical Court of Massachusetts, who consulted Prof. Michael Grodin of the Elie Wiesel Center for Jewish Studies at Boston University. Their opinion became known as the Vienna Protocol, due to the origins of the Pernkopf atlas. Their response may be found in this link, which I believe is better read in the full context of the Vienna Protocol than summarized by your humble blogger. For those of you who must read the opinion before reading the entire protocol, please follow the link and scroll to the 4th to last page at number 12 in section C entitled “The Protocol and Recommendations”.

The evil that created the Pernkopf atlas was the Nazi occupation of and executions that occurred in Austria during World War II. It is no longer occurring. No one in the present is suggesting that we resume executing people to gain more anatomic drawings to complete additional volumes of the atlas. Any good resulting from the current use of the atlas isn’t being offset by any ongoing evil of creating more atlas. The evil of the Pernkopf atlas is contained in the past and, in that sense, finite. Containing the evil seems to be a necessary step in obtaining good from that evil.

I mention this in closing as I believe there are current analogies of activities performed in the name of scientific good where we condone ongoing evil. Studying fertilized ova until sacrificing them on Day 14 (an evil) in the name of learning about human reproduction (a good) is one modern day example. In Vitro Fertilization done to obtain a healthy baby with genetic traits we want (a good) that results in the death(s) of other fertilized eggs we don’t want (an evil) is another. There are other examples we have discussed within this blog. I believe we need to contain and hopefully discontinue these and other practices if we want to claim the information we gain can honestly be called good.

Technical steps to gene-edited babies

This blog has carried several comments about the prospect of heritable human gene editing.  While nearly no one currently supports bringing such babies to birth—and condemns those who would rush ahead to do so—it appears a distinct minority think that we the human race should, if we could, agree never to do such a thing.  The most cautious perspective is to advocate a moratorium.  Others in favor of proceeding argue that, in essence, with the technologic genie (my term, not necessarily theirs) out of the box, a moratorium, much less a ban, is futile; the “rogues” will press ahead, casting off restraint. 

Advocates of research in this area have argued that a clear, careful, regulated pathway is needed to guide the work through necessary laboratory experiments that should be done first, before making a woman pregnant with a gene-edited embryo, in an attempt to be sure that the process is safe and highly likely to yield the intended result.  Even a moratorium would be, by definition, temporary, leaving the question, “when we will know to remove the moratorium?” to be answered.

A feature article in Nature, accessible without a paid subscription, asks “When will the world be ready” for gene-edited babies.  It walks through scientists’ understanding of what the technical issues are.  It is longer than a blog post, so I can only list key points here.  It is worth a reading by anyone interested, and it is written in sufficiently non-technical language that it’s accessible to the general, non-scientist public.

Key concerns are:

  • How would we be sure that genes that were NOT intended to be edited, in fact were not?
  • How would we be sure that genes that ARE intended to be edited are edited correctly?

These two matters have been addressed to some degree, or could be, in animals, but that would be faster and easier than in human egg cells or human embryos, and the results in animals may be different from what is found in the embryos.  (A further question is how many embryos, observed for how long, would need to be studied to support confidence.)

  • Even if the intended gene edit is made, is it clear that doing so is safe and does not induce other health risks? 

This blog recently reported the UK study that suggested that changes in the gene edited in the twin girls born in China last year might eventually reduce life span.  A criterion promulgated in 2017 by the National Academies of Sciences, Engineering, and Medicine was that the edited gene should be common in the population and carry no known risk (including, presumably, no increased risk) of disease.  Such knowledge is lacking for human populations, and what is believed known about the association of genes with risk of future disease has often been developed in Western populations, and may not apply to, for example, Africans.

  • At least some embryos would include some edited and some non-edited cells.  It would not easily be possible, if possible at all, to tell how many of which were present, or needed to be for the editing to work and not cause risks to the embryo’s development into a baby and beyond.  And what answers were obtained would require manipulating healthy embryos after in vitro fertilization.  The outcomes could not be predicted from first principles.
  • What should a clinical trial look like?  How many edited children would have to be born, and their health (and, most likely, the health of their progeny) observed for how long to get provisional answers before practicing the technique more widely?  Or, would the work proceed as IVF did—with dissemination in the general public, and no regulated research?

A US and UK committee is planned to address these questions, with the intent of proposing guidelines in 2020.  This will be important to follow, but with no chance to affect.  Most of us will just be watching, which leads to the last concern:

  • Is the world ready?

If that means, is there an international, or even a national, consensus, then the answer is clearly “no.”  That almost certainly remains “no” if one asks whether there is a future prospect for consensus.  It’s hard to envision something other than different groups and nations holding different judgments, and, most likely, remaining in some degree of irresolvable conflict.

Are AI Ethics Unique to AI?

A recent article in Forbes.com by Cansu Canca entitled “A New Model for AI Ethics in R&D” has me wondering whether the ethics needed for the field of Artificial Intelligence (AI) requires some new method or model of thinking about the bioethics related to that discipline. The author, a principal in the consulting company AI Ethics Lab, implies that there might be. She believes that the traditional “Ethics Oversight and Compliance Review Boards”, which emerged as a response to the biomedical scandals of World War II and continue in her view to emphasize a heavy-handed, top-down, authoritative control over ethical decisions in biomedical research, leave AI researchers effectively out-of-the-ethical-decision-making loop.

In support of her argument, she cites the recent working document of AI Ethics Guidelines by the European Commission’s High-Level Expert Group on Artificial Intelligence (AI HLEG). AI HLEG essentially distilled their AI ethical guidelines down to the familiar: Respect for Autonomy, Beneficence, Non-Maleficence, and Justice, as well as one new principle: Explicability. She downplays Explicability as simply the means to realize the other four principles. I think the demand for Explicability is interesting in its own right and will comment on that below.

Canca sees the AI HLEG guidelines as simply a rehash of the same principles of bioethics available to current bioethics review boards, which, in her view, are limited in that they provide no guidance for such a board when one principle conflicts with another. She is also frustrated that the ethical path researchers are permitted continues to be determined by an external governing board, implying that “researchers cannot be trusted and…focuses solely on blocking what the boards consider to be unethical.” She wants a more collaborative interaction between researchers and ethicists (and presumably a review board) and outlines how her company would go about achieving that end.

Faulting the “Principles of Biomedical Ethics” for failing to be determinant on how to resolve conflicts between the four principles is certainly not a problem unique to AI. In fact, Beauchamp and Childress repeatedly explicitly pointed out that the principles cannot be independently determinant on these types of inter-principle conflicts. This applies to every field in biomedical ethics.

Having an authoritative, separate ethical review board was indeed developed, at least in part, because at least some individual biomedical researchers in the past were untrustworthy. Some still are. We have no further to look than the recent Chinese researcher He Jiankui, who allegedly created and brought to term the first genetically edited twins. Even top-down, authoritative oversight failed here.

I do think Canca is correct in trying to educate both the researchers and their companies about bioethics in general and any specific bioethical issues involved in a particular research effort. Any effort to openly identify bioethical issues and frankly discuss potential bioethical conflicts at the outset should be encouraged.

Finally, the issue of Explicability related to AI has come up in this blog previously. Using the example of programming a driverless car, we want to know, explicitly, how the AI controlling that car is going to make decisions, particularly if it must decide how to steer the car in a no-win situation that will result in the death of either occupants inside the car or bystanders on the street. What we are really asking is: “What ethical parameters/decisions/guidelines were used by the programmers to decide who lives and who dies?” I imagine we want this spelled-out explicitly in AI because, by their nature, AI systems are so complex that the man on the Clapham omnibus (as well as the bioethicist sitting next to him) has no ability to determine these insights independently.

Come to think about it, Explicability should also be demanded in non-AI bioethical decision-making for much the same reason.