Preamble
In 2025, the United Kingdom implemented interlocking legal, policy, and administrative changes that significantly undermined the effective enjoyment of rights by transitioned women, despite the continued formal validity of legal gender recognition. No statutory right was explicitly abolished. Yet across core domains: safety, health, dignity, and participation – the practical capacity to access protection, services, and redress was systematically diminished.
This reconfiguration followed a cascading pattern. Threat-based public discourse contributed to a climate of danger (often described as stochastic terrorism). Institutional refusal to recognise testimony as valid knowledge ensured that resulting harm remained unrecorded and unaddressed (epistemic violence). These conditions, in turn, enabled the axiomatic prioritisation of natal sex over current material reality to be encoded into institutional procedures. Each layer reinforced the next, producing foreseeable and compounding harm that systems are structurally unable to detect, because they are not designed to see it.
The following developments did not occur in isolation. Taken together, they enacted a new administrative logic in which birth-assigned sex displaced lived identity and legal status as the operative criterion in systems mandated to ensure safety, care, and accountability.
Supreme Court Ruling (16 April 2025): The Supreme Court held that “sex,” “man,” and “woman” in the Equality Act 2010 refer to sex assigned at birth, and that a Gender Recognition Certificate does not alter this interpretation for sex-based exceptions. While the judgment did not mandate specific service rules, it was rapidly operationalised by public authorities as determinative. Legal recognition was preserved in form but decoupled from its protective function.
Policing and Safeguarding Practice: New national operational guidance directed police and safeguarding bodies to base detention, search, and placement decisions exclusively on birth-assigned sex, irrespective of anatomy, medical transition, legal status, or observable gender presentation. This approach exposed transitioned women to foreseeable risks of humiliation, physical danger, and psychological harm.
Data Collection and Statistical Invisibility: Revised guidance for official statistics removed gender identity as a standard variable, restricting sex recording to birth-assigned or legal sex without disaggregation. This change impeded the ability to identify, monitor, and address disparities affecting transitioned women. When harms cannot be measured, they cannot be meaningfully mitigated.
Expansion of Equality Act Exemptions: Throughout 2025, public authorities increasingly relied on statutory exceptions not as narrowly tailored, proportionate measures grounded in specific risk, but as presumptive bases for exclusion. Domestic abuse services, healthcare pathways, and community support spaces adopted blanket birth-sex criteria without individualised assessment or demonstrated necessity.
Media Language Shift: Major public service broadcasters and news organisations adopted “biological male” as the default descriptor for trans women, regardless of legal recognition, medical transition, or social integration. This language choice aligned public discourse with administrative practice, normalising exclusion and reinforcing stigma.
International Human Rights Responses: During 2025, international oversight bodies issued formal expressions of concern regarding the cumulative impact of UK policy and practice, identifying patterns of exclusion, loss of legal certainty, and heightened exposure to degrading treatment. A consistent theme was interdependence: when recognition is rendered ineffective, access to health, safety, participation, and dignity is predictably compromised.
These developments represent more than administrative drift. They reflect a coherent shift that preserves the form of rights while withdrawing their substance. Transitioned women are not denied recognition outright; they are repositioned from rights-holders entitled to protection as women to subjects managed as potential risks within safeguarding and service-design frameworks.
This undermines a foundational principle of human rights governance: that individuals must be treated in accordance with their lived, embodied, and socially recognised reality, not reduced to a single immutable datum that contradicts their legal and material existence.
This manifesto does not seek to re-litigate identity. It addresses a failure of governance. Systems designed to protect have been reconfigured in ways that no longer reflect the people they act upon, generating foreseeable and preventable harm to safety, dignity, health, and participation.
What follows sets out a response grounded in positive obligations and practical repair. It does not demand new rights. It seeks to restore the effective enjoyment of those already affirmed by redesigning systems to function accurately and fairly. It asks how institutions can meet their duties without forcing disclosure, requiring discretionary exceptions, or placing the burden of proof on those most exposed to harm.
The aim is functional justice: systems that classify not to sort, but to serve.
(I) Purpose
This manifesto sets out a practical approach to reducing preventable harm to transitioned women through targeted, evidence-informed change.
It is written for institutions, professionals, regulators, and members of the public who wish to act constructively. It does not require ideological agreement. It does not propose new rights. It focuses on restoring the effective enjoyment of existing protections by ensuring that systems function as intended in real-world conditions.
Our approach prioritises areas where harm is severe, foreseeable, and easiest to remedy through correspondence, guidance, and procedural change.
(II) Scope
This manifesto is concerned with the treatment of transitioned women whose transition is medically complete or clinically stable, and who live stably in their affirmed gender.
This focus is deliberate.
In this group, the gap between administrative classification and material reality is widest, and the consequences of misclassification are most severe. Bodies, physiology, medical risk profiles, and social roles have already changed in ways that are relevant to safety, healthcare, and safeguarding. Where systems continue to rely on birth-assigned sex as a default proxy in these circumstances, error is predictable and harm is foreseeable.
This manifesto addresses system design, not identity. It examines how institutions classify, route, and respond to people in safety-critical and care-critical contexts, and where those processes fail to reflect the reality they are meant to manage.
The scope of this work is intentionally limited to areas where:
- harm is acute rather than abstract,
- correction is possible without legislative change, and
- improvement can be achieved through guidance, correspondence, or procedural revision.
It does not attempt to resolve all questions relating to sex, gender, or identity, nor does it speak for all trans people. Different stages of transition involve different risks and needs, and no single framework can address them all responsibly.
This manifesto also assumes non-disclosure as a baseline condition. Many transitioned women cannot safely explain their history in moments of crisis or care. Systems that require disclosure in order to function fail those most exposed to harm. The scope of this work, therefore, includes only solutions that protect people without demanding visibility, exception-seeking, or personal explanation.
Finally, this manifesto is concerned with practical effect, not formal status. It asks whether recognition, once granted, continues to function in the places that matter: when someone seeks help, needs care, or requires protection. Where it does not, that failure falls within scope.
Everything that follows should be read with these limits in mind.
(III) Priority Domains
This manifesto prioritises a small number of domains where misclassification causes the most severe and immediate harm, and where correction is achievable without legislative change.
These domains are not selected because they are controversial, but because they are safety-critical, avoidable, and responsive to guidance and correspondence.
1. Policing, Custody, and Safeguarding
Decisions made by police and safeguarding bodies often occur at moments of acute vulnerability. Errors here expose people to immediate risk: humiliation, retraumatisation, physical danger, or disengagement from protection altogether.
For transitioned women with stable or complete transition, reliance on birth-assigned sex as a default proxy frequently conflicts with observable reality and clinical history. When rigid classification overrides context, professional judgment, or material conditions, the result is predictable harm.
This domain is prioritised because:
- consequences are immediate and severe,
- decisions are operational rather than legislative,
- discretion already exists in comparable safeguarding contexts.
Focus of action: Operational guidance, custody and search protocols, and safeguarding decision criteria where birth-assigned sex is treated as decisive by default.
2. Non-Gender Healthcare Pathways
Most healthcare interactions involving transitioned women are not about transition itself. They concern emergency care, inpatient placement, routine screening, and treatment for unrelated conditions.
Where administrative classification fails to reflect current anatomy, physiology, or risk profile, care is delayed, misrouted, or avoided altogether. For some, fear of inappropriate placement or treatment becomes a barrier to seeking help.
This domain is prioritised because:
- clinical reality is already well documented,
- misclassification undermines patient safety,
- routing and placement rules are amendable through guidance.
Focus of action: Ward placement and care-routing practices that default to birth-assigned sex rather than current clinical need, and harm generated through misclassification of physical and physiological reality.
3. Official Language and Administrative Classification
The language used by institutions does not merely describe practice; it shapes it. Descriptors that anchor people to historical classifications influence downstream decision-making, public understanding, and policy interpretation.
Where official forms, correspondence, or public communications use materially misleading descriptors, they normalise exclusion and restrict the range of responses considered legitimate.
This domain is prioritised because:
- language choices are discretionary,
- corrections do not require structural reform,
- impact cascades across multiple systems.
Focus of action: Terminology in official communications, forms, guidance documents, and public-facing reporting where classification no longer reflects material reality.
4. Safeguarding and Protection Services (Domestic Violence, Sexual Violence, Crisis Support)
Transitioned women experiencing domestic or sexual violence require safeguarding responses appropriate to the psychological context of abuse, trauma presentation, and barriers to safety they face. Where service design and referral pathways fail to pay due regard to these factors, support becomes ineffective or actively harmful.
Effective domestic violence and sexual violence support requires matching service design to the actual psychological context and safeguarding needs of the individual. Misrouting based on rigid birth-sex classification produces predictable outcomes: referral to services structured around different psychological contexts and trauma responses, disengagement from support altogether, or compounding of trauma through inappropriate placement.
This domain is prioritised because:
- service effectiveness depends on appropriate matching to the actual psychological context and need
- misrouting has immediate psychological and safety consequences
- referral pathways and service frameworks are amendable through guidance and training
Focus of action: Service design and referral pathways that pay due regard to the psychological context of abuse, trauma presentation, and actual safeguarding needs when routing transitioned women to support, rather than defaulting to birth-sex classification that produces systematic misreferral and disengagement.
(IV) Harm in Practice: Experience and Mechanism
The harms addressed in this manifesto arise where institutional power acts on a body it has chosen not to recognise as it is. They are not caused by malice or misunderstanding, but by specific mechanisms introduced or reinforced in 2025: rigid classification, procedural default, and the displacement of professional judgment.
For transitioned women, these mechanisms translate directly into fear, degradation, and withdrawal from protection.
1. Policing, Custody, and Safeguarding
Embodied harm produced by rigid classification
To be searched, detained, or supervised by the police is to be placed in a position of acute situational vulnerability – arising from the intersection of bodily exposure, institutional authority, and lived exposure to gendered violence. For a transitioned woman, that vulnerability is already shaped by the lived experience of male violence, sexual threat, and social scrutiny.
Being searched by male officers when one’s body is materially female is experienced as degrading and terrifying. It involves forced exposure, loss of bodily autonomy, and the knowledge that refusal carries consequences. For many, it echoes sexual assault. For others, it creates trauma where none existed before.
This harm is produced by a clear mechanism: classification by birth-assigned sex treated as decisive, regardless of current anatomy or lived role. Procedure overrides perception. Authority overrides reality.
For women who are stealth, the mechanism compounds harm. Compliance with procedure alone can disclose private medical history to strangers in positions of power. Disclosure is not chosen; it is extracted by design.
The outcome is foreseeable. When protection is delivered through degradation, people disengage. Complaints are withdrawn. Violence goes unreported. Safety systems fail precisely those who need them most.
2. Healthcare Outside Transition Services
Clinical vulnerability amplified by procedural routing
Healthcare requires trust. Patients submit their bodies to examination, confinement, and treatment on the assumption that systems will minimise harm and respect vulnerability.
For transitioned women, misplacement or misclassification communicates the opposite. Being placed in male wards, denied privacy, or treated as an administrative anomaly signals that one’s body is not being recognised as it is.
The mechanism here is procedural: routing and placement rules that prioritise birth-based classification over current anatomy, physiology, and risk profile.
In mental health settings, the effect is immediate. Distress increases. Agitation rises. Engagement drops. These outcomes are not controversial; they are basic psychiatric knowledge. When a patient feels unsafe in their environment, treatment efficacy declines.
In emergency medicine, the mechanism operates under pressure. Records that foreground birth sex steer clinicians toward assumptions that no longer fit the patient in front of them. Under time constraints, inaccurate proxies increase error.
The response is rational. Many women delay care, avoid admission, or endure symptoms rather than risk humiliation or exposure. The resulting harm is cumulative and predictable: later presentation, worse outcomes, avoidable suffering.
3. Administrative Classification and Official Language
Some of the most damaging harms do not occur in crisis. They occur through routine administration.
For transitioned women living stably post-transition, systems that require birth-based documentation or use descriptors that contradict lived reality create ongoing risk of involuntary disclosure. A form, a record mismatch, or an official letter can expose private history without warning or consent.
The mechanism is structural: administrative systems that treat disclosure as neutral and misclassification as inconsequential.
The consequences are well understood. Exposure can trigger job loss, housing instability, social hostility, or isolation. These outcomes do not require prejudice. They follow from how power operates when private information is revealed.
Language reinforces this mechanism. When institutions routinely describe transitioned women in terms that negate their lived reality, they normalise differential treatment. Description becomes justification. Justification becomes practice.
This is how bureaucratic language produces material harm.
4. Protection Services: When Service Design Fails to Match Need
For transitioned women experiencing domestic or sexual violence, harm occurs not only through categorical exclusion but through systematic mismanagement by services tailored to different abuse contexts and psychological presentations.
The mechanism is structural: services route people based on birth-assigned sex rather than the nature of abuse experienced, the emotional and psychological context of trauma, or actual safeguarding requirements.
Services addressing predominantly male survivors seem to focus on addressing externalising responses and accountability frameworks. Services addressing predominantly female survivors typically centre trauma-informed support for internalising responses – fear, shame, complex PTSD, survival strategies in contexts of coercive control, barriers to leaving created by emotional manipulation and economic dependence.
These patterns are not universal, nor exclusive to any sex or gender; they are well-established safeguarding indicators used across adult protection contexts wherever coercive control and trauma are present.
Variability in emotional presentation, including neurodivergent differences, does not negate vulnerability as a safeguarding factor; it alters how vulnerability presents and how support must be matched. Effective services match support to the actual psychological context and needs of the individual, not to categorical assumptions based on birth classification. Coping strategies and trauma responses vary across individuals, but where these patterns exist, appropriate support requires services equipped to address them.
When transitioned women are referred to services structured around different psychological contexts and trauma presentations, the result is often disengagement, not because support was offered, but because the support offered could not adequately address the psychological reality of the harm experienced.
The predictive failure compounds through specific vulnerabilities that transitioned women face:
- Perceived expendability: Some perpetrators treat transitioned women as though their emotional welfare matters less, their boundaries are less serious, and their needs are less legitimate. The reduced access to adequate safeguarding through misclassification compounds this.
- Reduced exit options: Absence of appropriate support services, belief that “no one else would want me,” or prior negative experiences with help-seeking create additional barriers to leaving
- Acclimatisation to harm: Chronic invalidation across multiple systems can normalise abusive treatment, making it harder to recognise danger or believe one deserves better
These vulnerabilities do not arise from birth classification. They arise from social positioning, lived experiences of relationships and vulnerability, prior encounters with rejection or invalidation, and institutional gaps in appropriate support.
The rational response to systematic misreferral is not to engage. Abuse continues. Isolation deepens. The service system records this as “lack of need” rather than “failure of design.”
This reflects established safeguarding knowledge: when people cannot access support appropriate to the psychological context and actual nature of harm they experience, systems fail in their protective function regardless of whether services formally exist. Effective safeguarding requires matching support to actual need, and avoiding categorical assumptions about birth classification. These distinctions already underpin risk assessment and service matching in adult safeguarding, domestic abuse, and trauma-informed care frameworks.
When Help Itself Becomes a Source of Harm
For people seeking help, systems are meant to reduce vulnerability, not intensify it. When institutional processes deny or contradict a person’s lived reality, the interaction itself can become harmful.
For transitioned women, this often occurs through repeated encounters in which:
- they are treated as deceptive, anomalous, or suspect rather than as ordinary service users,
- their bodies or identities are contradicted by authority figures acting under formal guidance,
- they are placed in environments experienced as unsafe or degrading.
These experiences are not isolated incidents. When they recur across police, healthcare, and administrative settings, they produce a cumulative effect. Each interaction reinforces the expectation that seeking help will involve humiliation, exposure, or loss of control.
The consequences are well established. Repeated institutional invalidation leads to:
- disengagement from services,
- delayed or avoided care,
- heightened anxiety and hypervigilance,
- long-term psychological harm.
This reflects known dynamics of trauma and power. When systems repeatedly negate a person’s reality at moments of vulnerability, they recreate the conditions of harm rather than alleviating them. When help is delivered through coercion, misrecognition, or forced exposure, it may be experienced as a repetition of power-over dynamics rather than protection, rather than a relief from harm.
Across all three domains, the same mechanisms recur:
- classification treated as permanent rather than predictive,
- procedure displacing judgment,
- disclosure assumed rather than protected against.
The harms that follow align with long-established knowledge about trauma, safeguarding, healthcare access, and institutional trust. The result is not only unmet need, but a predictable reduction in both engagement and effectiveness of systems designed to protect. When systems degrade people, they avoid them. When avoidance increases, harm deepens.
The Data Gap
Institutions correct harm by observing its effects. When those effects are not visible, systems cannot learn from their own failures.
For transitioned women, changes to data collection and classification practices in 2025 reduced the ability of institutions to see the consequences of their decisions. Experiences of misclassification, disengagement from services, delayed care, and exposure to harm were no longer consistently captured or distinguishable within administrative records.
The mechanism is straightforward. When categories flatten or erase lived experience:
- misclassification produces harm,
- that harm is not reliably recorded,
- lack of recorded impact is interpreted as a lack of problem,
- existing practices are reinforced rather than reviewed.
This creates a closed loop. Decisions generate harm, but the systems responsible are insulated from evidence of that harm. Absence of data is treated as reassurance, rather than as a warning sign.
Many of the harms described above do not appear immediately as complaints or incidents. They appear as absence: people who do not report violence, who do not attend appointments, who do not return after a degrading interaction. Without mechanisms to detect these absences, systems misinterpret withdrawal as resolution.
This is a failure of monitoring, not proof of effective policy. This is not merely a monitoring failure. It is data corrosivity: the systematic destruction of the evidence base through classification practices that render transitioned women statistically invisible. When categorisation collapses lived experience into birth-sex cohorts, specific harms – violence experienced as women, healthcare delays, safeguarding failures – are treated as statistical noise rather than detectable patterns.
This exemplifies epistemic violence: the refusal to recognise testimony and lived experience as valid sources of knowledge. Reports of harm are dismissed as “anecdotal”; absence of data is used to deny harm exists; and the burden of proof is placed on those already made invisible. The result is a self-sealing system that cannot learn from its own failures.
When harm cannot be seen, it cannot be addressed. When systems are designed to prevent visibility, the failure is architectural, not incidental.
(V) Once Harm Is Foreseeable, There Is a Duty to Act
The harms described in this manifesto do not arise from isolated incidents or individual misconduct. They arise from system design choices whose consequences are now visible, repeatable, and reasonably understood.
This section sets out what institutions owe people in those circumstances. It rests on basic principles of governance, safeguarding, and service provision that apply wherever public systems exercise power over vulnerable individuals.
1. The Duty to Prevent Foreseeable Harm
Institutions are not required to eliminate all risk. They are required to avoid predictable harm where reasonable alternatives exist.
Where evidence shows that a classification or procedure produces degradation, exposure, disengagement from protection, or avoidable injury, continuing that practice without review becomes a failure of care rather than an unfortunate side-effect.
In 2025, the harms arising from rigid classification, forced disclosure, and procedural override became foreseeable across policing, healthcare, and administrative systems. At that point, responsibility shifted from justification to prevention.
Systems must not continue practices that predictably cause harm simply because those practices are administratively convenient or historically entrenched.
2. The Duty to Enable Access to Protection
Protective systems exist to be used. When people avoid them because engagement itself causes harm, those systems are not functioning as intended.
Where procedures deter reporting of violence, delay access to care, or force individuals to choose between safety and dignity, access has been effectively denied even if services remain formally available.
Protection that depends on disclosure, exceptional treatment, or personal negotiation is not protection in any meaningful sense. It privileges those who can safely explain themselves and excludes those who cannot.
Systems must function for the people most exposed to harm, not only for those able to navigate complexity or tolerate humiliation.
3. The Duty to Ensure Dignity in State Interaction
Public institutions exercise authority. With that authority comes an obligation to minimise degradation, humiliation, and loss of bodily autonomy, especially in contexts involving confinement, examination, or enforcement.
When institutional processes repeatedly contradict a person’s lived and embodied reality, treat them as suspect or anomalous, or place them in environments experienced as unsafe, the interaction itself becomes harmful.
Dignity is not a symbolic concern. It is a functional one. People who experience state interaction as degrading disengage from systems intended to protect them, with predictable consequences.
Institutions must ensure that procedures designed to manage risk do not themselves become sources of harm.
4. The Duty to Monitor Impact and Learn From Error
Systems improve by observing their effects. Where harms are not visible, they are not corrected.
Changes to data collection and classification practices that obscure patterns of disengagement, delayed care, or exposure to harm undermine institutional learning. Absence of recorded impact cannot be treated as evidence of safety when withdrawal itself is a known response to degradation.
Monitoring is not a technical add-on. It is part of the duty of care. Without mechanisms to detect when people stop reporting, stop attending, or stop returning, systems misinterpret silence as success.
Institutions must ensure that their practices generate feedback capable of revealing harm, rather than concealing it.
5. Why Formal Recognition Without Effect Is Insufficient
Recognition that does not function in practice is not neutral. When legal or administrative status exists on paper but fails to protect in moments of vulnerability, the gap between form and effect becomes itself a source of harm.
In such circumstances, the issue is not the absence of recognition, but the failure to carry it through into operational decision-making.
Systems must be designed so that recognition, once granted, continues to have practical weight where it matters: when someone seeks help, requires care, or is subject to state power.
This failure reflects a deeper problem: when classification systems produce outcomes opposite to their stated purpose, they have failed on their own terms. For transitioned women, rigid natal-sex classification now consistently produces predictive failure (misrouting based on inaccurate proxies), procedural harm (rule-following that degrades), strategic disengagement (rational avoidance of harmful systems), and institutional blindness (data systems that cannot detect their own failures).
These are not edge cases requiring accommodation—they are systematic failures affecting thousands annually, requiring redesign. Once failures are foreseeable at this scale, continuing the system without fundamental review becomes a choice to accept harm rather than correct error. Institutions cannot claim neutrality when the consequences of their design choices are documented, repeated, and predictable.
6. Why Systems Must Work Without Requiring Disclosure
Many people cannot safely disclose intimate personal history in moments of crisis or care. Systems that assume disclosure as a prerequisite for safety shift responsibility away from institutions and onto individuals least able to carry it.
This is not a marginal concern. In the contexts described in this manifesto, forced disclosure is itself a source of harm.
Designing systems that work without requiring explanation, exception-seeking, or exposure is not an optional accommodation. It is a necessary condition of equitable protection.
From Foreseeability to Responsibility
Once harm is foreseeable, inaction is a choice.
The obligations described here reflect ordinary expectations of public systems entrusted with safety, care, and authority. They ask only that institutions attend to reality, correct predictable error, and prioritise material outcomes over administrative simplicity.
What follows sets out how these obligations can be met in practice.
(VI) Principles: How Systems Can Be Rebuilt to Work
The harms described in this manifesto arise from identifiable failures where a classification persistently misroutes people in safety-critical contexts. Across the priority domains, birth-assigned sex is being used as a classification proxy in contexts where it no longer reliably predicts risk, vulnerability, or care needs for transitioned women with stable, medically complete transitions.
Where a classification system produces systematically inaccurate predictions, its continued use degrades safety, displaces professional judgment, and generates foreseeable harm. In such circumstances, redesign is necessary for responsible governance.
In safeguarding and protection contexts, social exposure and psychological vulnerability function as predictive variables in the same way anatomy and physiology do in clinical contexts: they determine risk pathways, not identity.
This section sets out practical principles for rebuilding systems so that they predict accurately and function safely. These are operational standards grounded in comparative accuracy: where current biophysical state predicts outcomes more reliably than birth-assigned sex, systems must be designed to act on better information.
Each principle identifies a specific predictive failure, demonstrates why it occurs, and describes what accurate classification requires.
1. Classify to Predict, Not to Preserve
The predictive failure: Birth-assigned sex is used to predict risk, vulnerability, or clinical need in contexts where it no longer tracks the relevant variable. This is not a rejection of biological sex, but a rejection of its misapplication as a static proxy when dynamic indicators are available and more predictive.
Why prediction fails: For transitioned women with stable transitions, current anatomy, hormone-mediated physiology, and social vulnerability patterns have changed. Birth-assigned sex no longer predicts:
- Which custody environment minimises physical risk (current anatomy and vulnerability profile predict better)
- Which ward placement ensures patient safety and dignity (current anatomy predicts better)
- What screening or preventive care is clinically indicated (current physiology predicts better)
- What forms of violence someone is most vulnerable to (lived social role predicts better)
What accurate classification requires: Systems must prioritise classifications that predict the outcome they are designed to manage. Where birth-assigned sex fails as a predictor, continuing to privilege it over the current state reduces accuracy and increases error.
Classification exists to route people safely and appropriately. When it routes them inaccurately, it is failing its purpose.
2. Design Systems That Function Without Disclosure
The predictive failure: Systems assume that people will disclose transition history when relevant, failing to account for those who are stealth, who cannot or will not do so.
Why prediction fails: Many transitioned women live without disclosing their medical history. Systems that require disclosure to function correctly produce predictable outcomes:
- Misrouting in healthcare (the patient cannot safely explain why the male ward is inappropriate)
- Inappropriate search procedures (a person cannot explain anatomy without forced exposure)
- Data invisibility (harms occur but cannot be reported without unsafe identification)
What accurate classification requires: Systems must function correctly by default, using observable and recorded current-state information rather than depending on voluntary disclosure in moments of crisis or vulnerability.
A classification system that requires people to override it in order to be treated safely is not functioning accurately.
3. Assess Risk Using Current Indicators, Not Historical Proxies
The predictive failure: Birth-assigned sex is used as a blanket proxy for risk in contexts where it produces both false positives and false negatives.
Why prediction fails:
- False positives: Birth-assigned sex flags transitioned women as “male risks” in female spaces, despite evidence (e.g., Home Office 2021, HMICFRS 2023) showing they pose comparable risk to other women in those contexts
- False negatives: Birth-assigned sex fails to identify vulnerability to male violence, sexual assault, and harassment that transitioned women experience at rates comparable to or exceeding other women
What accurate classification requires: Risk assessment must use indicators that reflect current circumstances: patterns of vulnerability, documented history, observable presentation, and context-specific factors. Where blanket rules produce systematic misclassification, individualised assessment improves both safety and accuracy.
Accurate risk assessment distinguishes between people who present similar risk profiles and different risk profiles. Birth-assigned sex increasingly fails to do this for transitioned women with stable transitions.
4. Monitor Outcomes Without Requiring Exposure
The predictive failure: Data systems cannot detect patterns of harm, disengagement, or misclassification because they lack categories that distinguish transitioned women or because recording requires unsafe disclosure.
Why prediction fails: When harms are invisible to monitoring systems, institutions cannot learn from error. Absence of recorded complaints or incidents is treated as evidence of safety, when it may indicate:
- Avoidance of services due to prior degrading treatment
- Withdrawal of complaints due to forced disclosure requirements
- Inability to report without unsafe identification
What accurate classification requires: Monitoring systems must be able to detect patterns—disengagement from services, delayed care-seeking, adverse outcomes—using proxy indicators, aggregated data, and qualitative evidence that do not require individual identification.
Systems improve by observing their effects. Data practices that prevent observation prevent correction.
5. Enable Professional Judgment Using Accurate Information
The predictive failure: Rigid classification rules prevent professionals from acting on information they can directly observe, forcing them to ignore material reality in favour of administrative records.
Why prediction fails: Professionals closest to the situation—clinicians, custody officers, safeguarding workers—can often assess current anatomy, presentation, vulnerability, and risk more accurately than birth-assigned sex records indicate. When procedural rules override this assessment, accuracy decreases:
- Clinicians route patients based on records that contradict observable physiology
- Custody officers assign searches based on records that contradict the current anatomy
- Safeguarding workers places people in environments they can assess as unsafe, but cannot override
What accurate classification requires: Guidance must enable professionals to exercise judgment informed by current, observable reality. Classification should support decision-making, not replace it with inflexible proxies that reduce accuracy.
Where professionals are prevented from acting on information that would improve outcomes, the system is designed for administrative certainty rather than functional accuracy.
Accuracy as the Basis for Change
Taken together, these principles shift the basis for classification from administrative permanence to predictive accuracy.
They do not ask systems to abandon classification or ignore biological differences. They ask systems to classify using information that actually predicts the outcomes those systems are designed to manage.
Where birth-assigned sex predicts poorly—and where current biophysical state, lived vulnerability, or professional assessment predicts better—continued reliance on birth-assigned sex is not neutral administration. It is a choice to accept predictable error and foreseeable harm rather than improve system accuracy.
The principles above describe what accurate classification looks like in practice. They apply wherever prediction matters: in safety, healthcare, risk assessment, and institutional care. They are grounded not in rights discourse or identity claims, but in the basic expectation that systems designed to protect people should be able to recognise them as they are.
(VII) Implementation: Practical Steps for Change
This section translates principles into action. Each recommendation identifies the specific change needed and who has the authority to implement it.
Domain 1: Policing, Custody, and Safeguarding
What needs to change:
- Custody assessment protocols must include evaluation of current physical characteristics relevant to safe detention and search, not only recorded sex at birth
- Search procedures must prioritise officer assignment based on current anatomy and dignity considerations, with discretion to override default birth-sex assignments where this would cause foreseeable harm
- Safeguarding frameworks must recognise that rigid classification can itself create vulnerability
Who can implement this:
- College of Policing: Revise custody and search guidance to require assessment of the current physical presentation alongside administrative records
- Individual police forces: Adopt local protocols enabling custody sergeants to exercise judgment in placement and search decisions
- Police and Crime Commissioners: Include dignity in custody as a performance indicator
Domain 2: Non-Gender Healthcare
What needs to change:
- Ward placement protocols must prioritise current anatomy, clinical need, and patient safety over administrative sex markers where these conflict
- Clinical records systems must enable nuanced recording that reflects current physiology without requiring exposure of private history in every interaction
- Screening and prevention pathways must route patients based on current anatomy and hormone-mediated risk profiles, not birth-assigned sex
Who can implement this:
- NHS England: Issue guidance clarifying that ward placement should prioritise patient safety, dignity, and current clinical presentation
- Individual NHS trusts: Adopt placement protocols enabling clinical teams to override administrative defaults where this serves patient safety
- Royal Colleges: Incorporate principles of appropriate care routing in clinical education and standards
Domain 3: Official Language and Administrative Classification
What needs to change:
- Official communications must use descriptors that reflect the current reality, where this is relevant to the purpose of the communication
- Administrative forms must enable accurate self-identification without forcing disclosure of irrelevant medical history
- Public-facing language must distinguish between administrative data requirements and an accurate description of people as they are
Who can implement this:
- Individual institutions: Revise standard letter templates and form language to avoid materially misleading descriptors
- Professional bodies: Issue guidance on accurate, respectful terminology in official communications
- Ofcom and press regulators: Consider the accuracy implications of descriptors that contradict observable reality
Domain 4: Safeguarding and Protection Services
What needs to change:
- Referral pathways must pay due regard to the psychological context of abuse, trauma presentation, and barriers to safety when routing individuals to domestic violence and sexual violence support services
- Service design must enable appropriate matching based on actual need rather than categorical birth-sex assignment
- Assessment frameworks must recognise specific vulnerabilities – including perceived expendability, reduced exit options, and acclimatisation to harm that may not align with birth-classification assumptions
- Services must function without requiring disclosure that itself creates additional risk or compounds trauma
Who can implement this:
- Domestic Abuse Commissioner: Issue guidance emphasising appropriate service matching based on psychological context and actual safeguarding needs rather than categorical birth-sex routing
- Individual service providers: Adopt assessment protocols that evaluate the nature of abuse experienced, trauma presentation, and barriers to safety when determining appropriate support pathways
- Local safeguarding partnerships: Include protection-routing failures and service mismatch in safeguarding reviews and commissioning frameworks
- Professional bodies (British Association of Social Workers, College of Policing, Safeguarding): Incorporate principles of appropriate service matching in safeguarding training and standards
Conclusion
The harms described in this manifesto arise from identifiable system failures. They are not inevitable consequences of difficult tensions, but predictable outcomes of specific design choices.
Where classification no longer tracks reality, continuing to privilege it over current conditions is not neutral administration. It is a choice that transfers harm onto those least able to absorb it.
The principles and implementation steps set out here offer a different choice. They do not require abandoning safeguarding, creating special categories, or resolving complex ideological disputes. They require only that systems attend to the people they act upon, correct predictable errors, and function as intended.
This is work that can begin immediately. It does not require waiting for political consensus, legislative change, or resolution of broader debates. It requires recognising harm, understanding the mechanism, and acting to prevent foreseeable injury where reasonable alternatives exist.
For transitioned women, the stakes are immediate: whether seeking help means risking degradation, whether accessing care requires exposure, and whether protection systems function for those who need them most.
For institutions, the question is simpler: whether systems designed to protect are actually doing so, or whether they have been recalibrated in ways that reliably produce the opposite effect.
When harm is foreseeable, inaction is a choice. This manifesto sets out another path.
Edit: 3/1/25 We added a fourth domain: safeguarding and protection services. While this widens the remit, the understanding of transitioned women’s social, emotional and psychological framework is essential, and adds salient detail applicable to all other domains.