This is the first in a trial blog series called “Practically Prompted” – an experiment in using large language models to independently select a recent, ethically rich news story and then write a Practical Ethics blog-style post about it. The text below is the model’s work, followed by some light human commentary. See this post for the motivations behind this experiment.
The raw output and prompt used can be accessed here.
Below is the prompted post.
Should We Screen the Womb? Ethical Questions Raised by the New Miscarriage-Risk Test
Last Thursday researchers at the University of Warwick and University Hospitals Coventry & Warwickshire unveiled a laboratory test that analyses the molecular “decidual reaction” of the womb lining and predicts a woman’s risk of miscarriage before she conceives.¹ Drawing on more than 1,300 biopsies, the team argues that a persistently under-prepared endometrium explains many otherwise “unexplained” pregnancy losses and may be preventable. The NHS trust has already piloted the test with over 1,000 patients, and the authors hope it will move swiftly into wider clinical use.²
On its face this looks like unalloyed good news. Miscarriage is common (one in six pregnancies) and still poorly served by medicine. A tool that turns opaque tragedy into tractable pathology seems tailor-made to reduce avoidable suffering. Yet new diagnostics seldom arrive without moral baggage. Before the test migrates from Coventry to clinics across the UK—and, inevitably, to private fertility markets abroad—we should pause to ask what it might do, not just for, but also to, prospective parents.
From diagnosis to screening
The developers present the assay as a personalised diagnostic offered to women who have already suffered losses. But scale and enthusiasm could quickly convert it into a population-level screening tool, especially if marketing emphasises “peace of mind” for first-time parents. Screening carries a distinctive ethical profile: because most users are healthy, the tolerable threshold for false positives is lower, and consent must reckon with the worry generated by an uncertain risk score. In miscarriage research, where therapeutic options remain experimental, the danger of knowing without helping looms large.
Psychology, blame, and the politics of pregnancy loss
Shifting the aetiological spotlight from embryo genetics to maternal tissue may help some couples process grief, yet it also risks re-pathologising the pregnant body. Women already navigate a moral minefield of behavioural injunctions—avoid caffeine, alcohol, stress—under the implicit threat of blame should something go wrong. A test that labels the womb itself “defective” could reinforce the pernicious narrative that miscarriage is the mother’s failure, rather than a complex biological event. Robust counselling and public communication will be essential to counteract that tendency.
Justice in access and uptake
The Coventry pilot sits within the NHS, but if commissioning remains uncertain, well-resourced private clinics will fill the gap. Recurrent loss disproportionately affects women on lower incomes, who are also least able to pay out-of-pocket. Unless the test is equitably funded, it may widen health disparities—offering reassurance and bespoke therapies to the wealthy while poorer patients endure both higher risk and the stigma of “choosing” not to be tested.
Data, incentives, and unintended publics
Molecular read-outs of endometrial health will generate sensitive reproductive data. Even anonymised, such data may interest employers and insurers keen to quantify future pregnancy complications. Regulators must clarify that these scores remain strictly medical information, protected under data-protection law, with penalties for secondary use. Meanwhile, pharmaceutical firms will have incentives to develop “womb-priming” drugs positioned alongside the test; ethical review will have to keep pace with marketing claims that may outrun evidence.
Evidence, not just enthusiasm
The Science Advances paper underlying the test reports impressive correlations, but long-term clinical utility—actual reduction in miscarriage rates—remains to be demonstrated. NICE and other guideline bodies should insist on randomised trials before routine rollout, lest scarce maternity resources be diverted towards an intervention whose real-world benefits are unproven.
Conclusion
The new miscarriage-risk test exemplifies twenty-first-century precision medicine: molecular, personalised, and full of promise. It could spare thousands the anguish of repeated loss. It could equally medicalise normal uncertainty, exacerbate inequality, and deepen the moral surveillance of pregnancy. As clinicians, policymakers, and would-be parents weigh adoption, the central question is not whether we can predict the womb’s readiness, but how we ensure that doing so genuinely promotes reproductive flourishing.
Human commentary
It took o3 55 seconds to research and draft this post – though the bulk of its time was spent trying to adhere to the 600-word limit.
It is fascinating to look at its “Thoughts”. After identifying the past week’s dates, it considered several topics – the EU AI Act, CRISPR therapy, UK embryo research – before scanning the web and selecting the new miscarriage-risk test.
Why that topic? Who knows! But I think the choice was strong—the issue is salient, novel, and ethically dense, intersecting with consent, blame, equity, and data ethics. So far as I can tell, it reported on the news component of it accurately.
In terms of the analysis itself, it covers the obvious questions (medicalisation of uncertainty, psychological burden and blame, equity in access, consent and screening ethics, data privacy and misuse. commercialisation and regulatory lag) – impressive given the word limit constraints.
On the other hand, the breadth of issues covered meant that none of them were dealt with at any depth – a common trade-off in a short, news-based post. Nonetheless, I wanted more from its argument that a test identifying issues with the womb could reinforce the “pernicious narrative that miscarriage is the mother’s failure”. An alternative viewpoint, not explored, is that locating the issue in a treatable physiological condition could lessen feelings of personal failure and blame for some women. But maybe that’s too much to ask for in such a short post.
Ultimately, there’s clear evidence that many people are highly averse to closely reading and engaging with text that is labelled “AI”. Perhaps this is changing. Did you read this post differently knowing its author wasn’t human? What did you think of it? Let us know in the comments below.