Done carefully enough, it is not a crime

Advocacy — Conversion Practices Bill

The government’s draft Conversion Practices Bill makes conversion practice a crime only when it is abusive and has been proved to cause serious harm. Someone who is patient, gentle and careful can pursue the same aim, over the same months, and stay outside the offence unless serious harm can be proved to have followed. This piece explains why, and what would close the gap. We will be submitting written evidence to the Joint Committee of both Houses that is now scrutinising the Bill.

A conversion practice is a violation of the person it is done to, whether it is done roughly or carefully. It sets out to change who someone is, or to make them hide it, on the premise that who they are is a defect to be corrected. That premise is what makes it wrong. The draft Bill does not quite say so. It makes the practice a crime when it is done badly enough, and leaves it alone when it is done well.

We support the Bill

It is worth being plain about this first, because nothing that follows is an argument against it.

The Bill is trans-inclusive: it covers transgender identity as well as sexual orientation, and its definition expressly includes people who identify as neither male nor female. It creates an offence of encouraging or assisting a conversion practice carried out abroad on someone from the UK. It creates civil protection orders that a court can make before any harm is done, on the basis that a practice is likely to cause it. It reaches companies, charities and other organisations, and the people who run them. That architecture is sound, and it has taken years to reach. We want it tightened, and we want a tightened Bill passed.

Three gates

Under the Bill, conduct becomes a crime only if it passes three gates. It must be a conversion practice: conduct intended to cause a person to have, or not to have, a sexual orientation or transgender identity, or to believe that they do or do not have one (clause 1(2)). It must be abusive, judged on all the circumstances, with the court considering whether it involved sexual, violent, threatening, controlling or coercive behaviour, or economic, psychological or emotional pressure (clause 1(4) to (6)). And it must have caused serious harm to the person’s physical or mental health, or serious alarm or distress with a substantial adverse effect on their usual day-to-day activities (clause 2(1)).

The Government’s own Explanatory Notes say why the Bill is needed: existing law already covers violent acts, but “this is not the case for all conversion practices, particularly those which consist of verbal or non-violent physical acts” (paragraph 22). Those are the practices the Bill is for. They are also practices whose effects can be hard to prove to the standard the third gate demands.

The Bill punishes the harm, once it can be proved. It does not prohibit the practice.

What careful looks like

Take a paid “coach” who holds weekly sessions over a year with a sixteen-year-old, with the aim of making them stop identifying as trans. There are no threats. The sessions are warm, patient and persistent. The distress comes later, after the sessions have ended, alongside everything else happening in a teenager’s life.

Under the Bill as drafted, that is a crime only if a court finds the sessions abusive and is sure they caused serious harm. Harm from months of quiet pressure emerges slowly, sits alongside other causes, and is often documented only after the relationship has ended. We do not claim to know how often such prosecutions would fail. We say that the practices the Bill was written for present particular difficulties for an offence that turns on proving the result.

The Bill already accepts that risk can be enough. A person who helps arrange a practice abroad commits an offence if it is likely to cause serious harm, and a court can make a protection order on the same basis. We are asking for the same test for conduct here: an offence of abusive conduct that is likely to cause serious harm, which the person knew or ought to have known, alongside the Bill’s existing offence, which keeps its requirement of proven harm and its five-year maximum. The abuse requirement stays; conduct a court does not find abusive is not caught.

Hiding may not be covered

Conversion practice does not always dispute what a person is. It can instead require them to hide it: never to say it, never to act on it, never to live as who they are. The Bill’s definition covers conduct intended to make a person have, or not have, an orientation or identity. Pressure to hide can be the means to that end, and then it is covered. But where concealment is the only aim — the person may stay who they are, so long as nobody sees it — the definition may not apply, however the pressure is applied. Naming suppression expressly would remove that uncertainty.

The survey evidence cannot say how often concealment alone is the aim, but it shows how widely these experiences are reported. In a 2022 survey of 2,042 LGBT+ adults in the UK, published by the charity Galop in January 2023, 18% said someone had tried to change, “cure” or suppress their sexual orientation or gender identity — rising to 43% of binary trans respondents and 36% of non-binary and gender-diverse respondents. Those figures answer a single broad question, do not separate suppression from other forms, and are not estimates of conduct that would meet any criminal offence. But the laws of the Netherlands, Germany, New Zealand and Victoria name suppression expressly, and the laws of Malta and Canada reach it with the word “repress” (Canada’s in one direction only). The Bill does neither.

A clinic should not be the safest place to try

Clause 1(3) says that conduct carried out with a conversion intention, in the course of providing health care, is not a conversion practice at all unless the person acts in a way that “falls far below” the standards reasonably expected of them. The Explanatory Notes describe that as “equivalent to a threshold of gross negligence” (paragraph 31), and say the health care definition reaches people working in “regulated and non-regulated sectors” (paragraph 32).

Read carefully, that exclusion does not protect clinicians who explore a patient’s identity with an open mind: they have no conversion intention, so they were never within the definition. It protects someone who does intend to change a person’s orientation or identity, provided they do it in the course of health care and not grossly badly. Conversion practice should not be happening in clinical settings at all, and a clinical setting should not be the one place where intending it carries a gross-negligence threshold.

We are asking for a narrower, clearer rule: a safe harbour for registered or accredited practitioners, and those they supervise, whose care is clinically appropriate in their reasonable professional judgement and consistent with their register’s published standards — withdrawn wherever the practitioner pursues changing or suppressing who a person is as an objective imposed on them — together with provisions making clear, for everyone, that exploration, information-giving and support for a person’s own decisions do not by themselves show a conversion intention. A clinician who supports a patient’s own exploration and decisions keeps that protection. One who sets out to change who a patient is cannot claim it.

Children, and consent

The Bill has no separate rule for children. A practice aimed at a child must clear the same abusive-and-seriously-harmful bar as one aimed at an adult. Germany prohibits conversion treatment of under-18s, exempting parents unless they grossly breach their duty of care; New Zealand criminalises conversion practices on under-18s, and on people who lack decision-making capacity, without proof of harm; the Netherlands will criminalise systematic or intrusive practices in professional and organisational settings from 1 January 2027. Each strikes the balance differently, with different safeguards. We are asking for an offence covering professional, paid or organised conversion practice on a child, or on an adult who lacks capacity, directed at them because of their orientation or identity, where it is part of a planned, organised or sustained course of conduct or is otherwise intrusive — the Dutch threshold — without proof of abuse or harm, with a maximum of two years and with prosecutions requiring the consent of the Director of Public Prosecutions.

The Bill is also silent on consent. Someone who has been told for months that who they are is a sickness may well “agree” to be cured of it. We are asking for the Bill to say that apparent consent does not of itself stop conduct being abusive, that consent is no defence to the offences concerning children and adults who lack capacity, and that the person subjected to a practice can never be prosecuted for taking part in it.

What we are not asking for

We are not asking for every conversation about identity to become a crime. Prayer, worship, teaching and the general expression of belief stay lawful; so do a person’s own choices about whether, when and to whom to come out, and advice given to help them make those choices. A parent’s ordinary conduct stays outside the offences unless it is abusive. Clinicians practising to their professional standards keep a safe harbour. A single conversation would not be caught by the child offence unless it was intrusive in itself, and repetition alone would not be enough. The new offences need the consent of the Director of Public Prosecutions, and we are asking for statutory guidance to be published before the offences start, with a fixed date by which they start regardless.

Alongside those, we are asking for an offence of offering or advertising such practices in a professional, paid or organised setting, including programmes sold to parents for use on their children; for arranging a person’s travel abroad for a practice to count expressly as assisting it; and for a baseline, and a report to Parliament within three years on how the Act is working, including any unintended effects on health care, education and freedom of belief.

We have set out our four principal recommendations, and the limits built into each, on a separate page: what we are asking Parliament to change.

What happens next

A Joint Committee of both Houses was appointed in September 2026 to consider the draft Bill, and is due to report by 22 January 2027. As of 11 October 2026 it had not yet opened a call for evidence. We have told the Committee that we intend to submit written evidence, and we will do so when it opens. The evidence will set out the gaps above, with drafted amendments to the Bill’s text. We will publish it here once the Committee has published it.

A violation done carefully is still a violation. A law against conversion practice should be able to say so.

Statutory text checked against the Draft Conversion Practices Bill (CP 1604, GOV.UK, page updated 1 September 2026) and its Explanatory Notes, paragraphs 22, 31 and 32. Survey figures from Galop, There was nothing to fix (January 2023), printed page 6; the methodology is on pages 34 to 36. Eden Openly receives no funding from, and has no commercial relationship with, any provider of gender-related healthcare or any pharmaceutical company. The author is a transgender woman and declares her interests in full on our declared interests page; see also our statement of independence. Corrections and queries: [email protected].

Version history. v1.0, 10 October 2026 — first published. v1.1, 10 October 2026 — corrected after a further review, the same evening. The piece had said that conduct intended to make someone hide who they are is not covered by the Bill “at all”; where hiding is the means to changing someone, it may be, so the piece now says the gap is where concealment is the only aim, and that the definition may not apply. It had called suppression “common” and said much conversion practice seeks concealment; the survey cannot separate concealment from other forms, so those claims are removed. It had said the practices the Bill targets are the “least likely” to leave proof; that ranking is not established and is removed. Our proposals on health care and children have also been settled more precisely: a safe harbour for care given to professional standards, and a child offence limited to systematic or otherwise intrusive conduct. The piece now describes those. v1.2, 11 October 2026 — brought partly into line with our draft evidence: the health care safe harbour is tied to the practitioner’s register and withdrawn where an outcome is imposed on the patient; the child offence applies to conduct directed at a child because of their orientation or identity, with a two-year maximum. v1.3, 11 October 2026 — corrected, and aligned with our page on what we are asking for. Germany’s prohibition, which we had called “outright”, exempts parents unless they grossly breach their duty of care. Malta, which we had listed as naming suppression, uses the word “repress”, as does Canada; the list of laws now matches our asks page. Clause 2(1) refers to “usual day-to-day activities”, not “day-to-day life”. Galop’s methodology is on pages 34 to 36, not page 35 alone. The piece had said a single conversation would not meet the child offence; our proposal catches one that is intrusive in itself. Consent is no defence for adults lacking capacity as well as for children. The piece now sets out our fourth recommendation, on likely harm, and our supporting proposals on advertising, removal abroad and review. The summary no longer says careful conversion practice is “not a crime at all”. The v1.2 note above, which had said the piece was brought into line with our “final” draft evidence, now says “partly”, because at v1.2 the piece did not yet cover all of our recommendations. The status of the Committee’s call for evidence is now given as at 11 October 2026.