Online peer support and crisis text services

1. Introduction

Version followed: the services as they describe themselves on their own pages, checked October 2026: 988 Suicide & Crisis Lifeline Terms of Service (last updated August 2026), its police-intervention FAQ (last modified October 2024) and Suicide Safety Policy (revised June 2024); Crisis Text Line Privacy Policy (updated 17 April 2026); Shout "How Shout works" and Service User Privacy Notice (updated 29 June 2026); Samaritans contact page; Togetherall "About" FAQ (modified 24 August 2026); Orygen #chatsafe second edition (May 2023).

These are not a psychotherapy but a group of services that people can reach without a referral: crisis lines run over text, chat or messaging; and moderated peer communities where members support one another under staff oversight. A psychotherapist does not deliver them. The therapist's work is to know what each service offers and states about confidentiality and escalation, to help the client choose and prepare to use them, to place them in a written safety plan, to talk about the risks of unmoderated online spaces and pro-self-harm content, and to debrief afterwards. The underlying idea of change is simple: a person in acute distress who can reach a trained human quickly, by a channel they will actually use (often text rather than voice), can be helped to de-escalate and get through the moment, and a peer community can reduce isolation between sessions; the therapist's job is to make that link usable, safe and connected to ongoing care. [source: 988 Suicide Safety Policy; Shout How Shout works; Togetherall FAQ] [synthesis for the therapist's role]

None of these services replaces therapy, emergency services or a clinician's own crisis pathway. Policies differ between services and change; describe each one to clients in the service's own published terms, with the date checked, and avoid categorical promises about confidentiality or about whether police or ambulance will or will not be involved. [synthesis]

2. Framework

This is a framework for the therapist's work alongside the services, not a manualised protocol. [synthesis throughout this section, built on the services' published descriptions]

Phase A. Preparation (before you need it; clinician-level, ongoing). Goal: a current, checked list of services for the jurisdictions your clients live in, with each service's published confidentiality and escalation terms, opening hours, channels and age limits, plus your own service's crisis pathway. Move on when you can describe each listed service accurately in two or three sentences. Recheck at least every six months, since services close channels (for example, Samaritans in the UK states its email and letter services have closed and web chat is a small pilot) [source: Samaritans contact page].

Phase B. Assessment (first sessions; repeat when risk changes). Goal: understand the client's current suicide and self-harm risk using your service's standard risk-assessment practice; how they already seek help online (crisis lines used, forums, social media, AI chat); exposure to self-harm or suicide content; and preferences (voice versus text, anonymity, fears of police involvement). Move on when you know which channels the client would realistically use in a crisis.

Phase C. Collaborative planning (usually one session, part of safety planning). Goal: the client chooses one or two crisis services and, if wanted, one moderated peer space; these are written into the safety plan with numbers, opening hours and what to expect; the confidentiality and escalation terms each service publishes are discussed; and a first message is rehearsed. Move on when the client can say what they would do, in what order, and has the contacts stored where they can reach them.

Phase D. Use and debrief (whenever a contact happens). Goal: review any crisis contact or forum episode in the next session: what helped, what did not, any emergency response, how the client feels about using it again. Update the plan.

Phase E. Review and ending. Goal: the plan is reviewed at set points (after any crisis, at changes of risk, before breaks in therapy, and at discharge), and before ending the client has a plan that does not depend on you.

Typical format: integrated into ordinary individual sessions (in person or remote). Time needed: a full session for initial planning, then 5 to 15 minutes at reviews [synthesis]. For young people, safeguarding, consent and parent or carer involvement follow your service's and jurisdiction's rules (see 4.13).

3. Guiding principles

[synthesis unless a source is given]

Describe services in their own published terms, dated. Say "as of October 2026, their privacy policy says…" rather than "it's confidential" or "they never call the police".
No categorical claims about confidentiality or police involvement. Every service opened lists exceptions (imminent risk, attempts in progress, safeguarding of children and at-risk adults, legal requirements); explain these honestly so the client is not blindsided. [source: 988 Terms; Crisis Text Line Privacy Policy; Shout Privacy Notice; Samaritans contact page]
Least invasive response first is the stated aim of the main crisis lines. 988's policy requires collaborative safety planning and the least invasive intervention, with involuntary emergency response only after less invasive options fail or when an attempt is under way; Shout states it first tries to agree a safety plan. Use this accurately to reduce fear, without promising outcomes. [source: 988 Suicide Safety Policy; 988 police FAQ; Shout Privacy Notice]
Match the channel to the client. Text and chat suit people who cannot speak aloud safely, are deaf or hard of hearing, have social anxiety or are young; voice may suit others. Ask, do not assume.
Crisis lines are for the moment; peer forums are for between moments; therapy is the ongoing work. Keep the three roles distinct in the plan. Togetherall, for example, states it "is not a crisis service". [source: Togetherall FAQ]
Moderation matters. Prefer peer spaces that publish moderation, house rules and an escalation process; treat unmoderated forums and pro-self-harm content as risks to be discussed openly, not forbidden in a way that drives secrecy.
Safe communication online can be taught. Orygen's #chatsafe guidelines, developed with young people, exist so young people can talk about self-harm and suicide online more safely; use them rather than a blanket "don't talk about it". [source: Orygen #chatsafe pages]
Data is part of informed choice. Conversations are recorded and kept for set periods by at least some services (for example, up to seven years for identifiable data at Crisis Text Line and at Shout) and anonymised data may be used for research or training; clients deserve to know this. [source: Crisis Text Line Privacy Policy; Shout Privacy Notice]
The therapist stays responsible for their own clinical pathway. A crisis line in the plan does not replace your service's out-of-hours arrangements, your duty to act on risk you learn of, or supervision.
Curiosity over judgement about past experiences. Clients may have had a distressing contact (long wait, unhelpful responder, an emergency dispatch). Explore before recommending again.
Never give means information. In discussing harmful content, talk about its effect and how to avoid it, not about methods; lethal-means conversations stay at safe storage and removal.

4. Intervention inventory

Part A (4.1 to 4.6) describes the services as they publish themselves, so you can explain them accurately. Part B (4.7 to 4.18) is the therapist's own work. Status lines are dated; recheck before relying on them.

4.1 Service profile: 988 Suicide & Crisis Lifeline (US) call, text and chat

Purpose and when to use: US clients in suicidal, mental health or substance-use crisis who prefer, or can only use, text or chat. 988 went live nationwide on 16 July 2022, replacing a 10-digit number, and is reachable by call, text or chat; it is a network of more than 200 state and local centres, administered by Vibrant Emotional Health with SAMHSA/HHS support. Spanish text and chat were added around the first anniversary (text AYUDA to 988, or chat at 988lineadevida.org). [source: APHA Public Health Newswire, July 2023] As of October 2026, a site banner states that the LGBTQ+ youth "Press 3" option has been restored [source: 988 police FAQ page]; confirm current availability before naming it to a client [uncertain: not verified].
Confidentiality as published (as of October 2026 per the 988 Terms of Service, last updated August 2026): information will not be shared outside the 988 Lifeline without consent, except in listed situations: imminent risk of physical harm, a suicide attempt or overdose in progress, possible medical emergencies, and abuse or neglect of a minor or elder as state law requires. Contacts are not required to give identifying information or a specific location, though in urgent cases help may be sent to their location. The Terms advise calling 911 first if there is an immediate physical threat.
Emergency escalation as published: the 988 FAQ states counsellors contact emergency services only when risk to self or others is imminent or in progress and a less invasive plan cannot be agreed; it lists alternatives (safety planning, mobile crisis teams, involving loved ones or professionals, crisis stabilisation units). It states fewer than 2% of 988 calls involve emergency services and over half of dispatches are with consent (these figures refer to calls). [source: 988 police FAQ, modified October 2024] The Suicide Safety Policy (effective December 2022, revised June 2024) requires centres to ask every contact about suicidal thoughts, use a safety assessment model, plan collaboratively, use the least invasive intervention, reserve involuntary intervention for when less invasive options fail or an attempt is under way (it may be with or without consent), have supervisors available at all hours with review of dispatches, and give chat/text-specific training. [source: 988 Suicide Safety Policy]
Steps (for explaining to a client): text or chat; a trained counsellor at a network centre responds; the counsellor will ask about suicidal thoughts; the aim is to help the person be safe and make a plan; emergency response is described as a last resort with stated exceptions.
Example language (illustrative): "If you text 988, a trained crisis counsellor answers. They will ask directly about suicide; their published policy has them ask every contact, so it's a routine question. Their published policy says they aim for the least intrusive way to keep you safe, and they list situations, like an attempt in progress, where they may send help."
Watch for: fear of police contact (explore, give the published figures and exceptions accurately); clients outside the US (988 is US-only); minors (see 4.13).

4.2 Service profile: Crisis Text Line (US) and the 2022 Loris.ai data-sharing controversy

Purpose and when to use: a US text-based crisis service (also via other messaging apps, which follow their own privacy policies). As of October 2026 its privacy policy states its crisis services are for the US only and refers people in the UK, Canada and Ireland to Shout, Kids Help Phone and Text About It. [source: Crisis Text Line Privacy Policy, updated 17 April 2026]
Data as published (as of October 2026 per the Privacy Policy): records of every message, phone number, carrier and IP address are collected; identifiable information is kept up to seven years, then anonymised or deleted; non-identifiable data is kept indefinitely and may be shared with researchers under contract and, in the policy's words, with third parties "for any reason". Texting STOP ends a conversation; DELETE requests deletion of personal data and transcripts. The organisation states it is a nonprofit exempt from COPPA, and the policy sets no minimum age or parental consent requirement.
Escalation as published: emergency services may be contacted when it believes this will prevent death or serious injury; the policy states that just saying you feel suicidal "does not necessarily mean" emergency services will be called, and that contact is more likely with intent, ability, a near-term plan and inability to de-escalate. It reports abuse or neglect of children under 18 and of disabled, dependent or elderly adults to authorities as required by law. [source: Crisis Text Line Privacy Policy]
The 2022 controversy as reported: in late January 2022 Politico reported that Crisis Text Line shared anonymised conversation data with Loris.ai, a for-profit spin-off (launched 2021, making customer-service training software) in which the nonprofit held a stake and which shared a founder, Nancy Lublin; privacy experts questioned whether users understood and whether data could be truly de-identified. Crisis Text Line initially defended the arrangement as public and said it shared no personally identifiable information. On 31 January 2022 FCC Commissioner Brendan Carr wrote asking it to end the arrangement; that evening it announced it had ended data-sharing with Loris, updated its terms and privacy policy and asked Loris to delete the data. [source: Popular Science, 1 February 2022, reporting Politico]
Steps [synthesis]: (1) Confirm the client is in the US. (2) Open the current Privacy Policy together and note its date. (3) Explain what it says about records, texting STOP and DELETE, and when emergency services may be contacted. (4) If the client raises the 2022 controversy, give the dated facts. (5) Record the choice and the date checked in the safety plan.
Example language (illustrative): "Crisis Text Line keeps records of conversations for a period and uses anonymised data for research. In 2022 it was criticised for sharing anonymised data with a related company and ended that. If that worries you, you can text DELETE afterwards to ask for your data to be removed, or we can look at other options."
Watch for: clients who have read about the controversy and distrust all text lines (give the dated facts, not reassurance); clients who need the service right now (don't let a data discussion delay crisis use).

4.3 Service profile: Shout 85258 (UK)

Purpose and when to use: a free, 24/7 UK text service for anyone struggling; text SHOUT to 85258; messages do not appear on the phone bill. A conversation starts with four automated messages, then a trained volunteer asks what is troubling the person; the aim is to reach a calm, safe state and a plan, with signposting to further support; conversations usually last 45 to 60 minutes. Its page advises calling 999 if life is at imminent risk. [source: Shout, How Shout works, checked October 2026]
Confidentiality, data and escalation as published (as of October 2026 per the Shout Service User Privacy Notice, updated 29 June 2026): run by the charity Mental Health Innovations. Responders cannot see the texter's number; message content and responder notes are kept up to seven years after last contact; anonymised aggregated data is kept indefinitely, shared with selected partners and used to train an AI "conversation simulator" for responder practice. Responders are described as "supervised throughout" by qualified mental health professionals. Details may be shared with police, ambulance or social services where there is imminent risk to life or a safeguarding concern; the service first tries to agree a safety plan, and if that is not possible a clinician may share details. Some technical staff in other countries, including the US and New Zealand, may access data.
Age: the notice does not state a minimum age; Shout links to a separate page for under-25s [uncertain: age policy not verified].
Steps [synthesis]: (1) Confirm the client is in the UK. (2) Describe how a conversation runs, as Shout describes it. (3) Read the privacy notice's escalation and data sections together, noting its date. (4) Make clear each contact is a single conversation, not ongoing support. (5) Add the number to the safety plan and the client's phone.
Example language (illustrative): "Shout is text only, free, and doesn't show on the bill. A volunteer, backed by clinical supervisors, will text with you for maybe 45 minutes to an hour. If they think your life is at immediate risk, they say they'll try to make a plan with you first, but they may share your details with emergency services."
Watch for: clients expecting ongoing support (it is single conversations); clients outside the UK.

4.4 Service profile: Samaritans (UK and Ireland)

Purpose and when to use: 24/7 free listening support by phone on 116 123 "day or night, 365 days a year"; Welsh Language Line; Relay UK for d/Deaf people and those with hearing or speech impairment; in person at some branches. As of October 2026 the contact page states the UK email service and letter service have closed and web chat is a small-scale pilot "not currently available" to everyone. So for UK clients who need text, Samaritans is not currently the main text option. [source: Samaritans contact page, checked October 2026]
Confidentiality as published: Samaritans states it maintains confidentiality, including after a caller's death, and lists exceptions: the caller agrees; an ambulance is called because the caller appears unable to make rational decisions; a court order or law requires it; information about terrorism, bomb threats or threats to life; attacks or threats on Samaritans staff or volunteers; deliberately blocking the line; harm or risk of harm to a child or at-risk adult (safeguarding policy); and disclosed plans that could seriously harm a third party, where it may contact emergency services with or without consent if it has identifying details. Numbers are not shown to volunteers and callers need not give a name. [source: Samaritans contact page]
Online safety work: in September 2020 Samaritans published guidelines for technology companies hosting user-generated content, on minimising access to harmful self-harm and suicide content and creating opportunities for support, alongside an Online Harms Advisory Service. Its release cites research that at least a quarter of young people presenting to hospital after self-harm with suicidal intent had used the internet in connection with the attempt. [source: Samaritans news release, 10 September 2020]
Steps [synthesis]: (1) Confirm the client is in the UK or Ireland. (2) Check the contact page for current channels before naming any (email and letters have closed; web chat is a limited pilot). (3) Go through the published confidentiality exceptions together. (4) If the client cannot or will not use the phone, offer a text option such as Shout. (5) Record the choice and the date checked in the plan.
Example language (illustrative): "Samaritans is a phone line, any hour. Their published policy says they keep what you say confidential, and it lists exceptions, for example if a child is at risk, or if they believe you can't make decisions for yourself and call an ambulance. Let's look at the current page together."
Watch for: clients who remember Samaritans email and expect it; Samaritans services outside the UK and Ireland are separate organisations with their own policies [synthesis].

4.5 Service profile: Togetherall (moderated peer community)

Purpose and when to use: an anonymous online peer community open 24/7 (formerly Big White Wall, renamed in 2020): members post in Talkabouts, join groups and one-to-one chats, make creative posts, take self-assessments and do self-help courses. Access is through commissioners (NHS, councils, insurers, employers, universities and colleges; free for UK service personnel, veterans and families through Ministry of Defence and NHS England partnerships); the page states there is no individual paid subscription. Minimum age 16. [source: Togetherall About FAQ (UK), modified 24 August 2026]
Moderation and escalation as published: "professionally trained Wall Guides" monitor the community, with all new posts monitored 24/7; they respond proactively to members in crisis and may help them contact a local doctor or therapist; specified behaviour triggers an escalation process; content can be removed and members barred for breaking house rules; members are advised not to exchange personal details. The page states it "is not a crisis service" and points to an In Crisis page. The page does not describe contacting emergency services or disclosing identity [uncertain: escalation details beyond this not verified; read the current House Rules and Privacy Policy]. Members can request deletion of personal data. [source: Togetherall About FAQ]
Steps [synthesis]: (1) Check eligibility through local commissioning and the published minimum age. (2) Describe the moderation and house rules as published, and that it states it is not a crisis service. (3) Agree with the client how they will use it (between times) and what goes in the safety plan for crises. (4) Review its effect on mood and urges in later sessions (4.16).
Example language (illustrative): "Togetherall is somewhere to talk to other people who get it, any time, anonymously, with trained staff watching the posts. It's for in-between times, not for emergencies; your safety plan covers those."
Watch for: eligibility (depends on local commissioning); over-reliance; distress from others' posts.

4.6 Other services to verify before naming

Purpose and when to use: when a client uses or asks about a service not profiled above. Services such as Kooth (young people, UK) and 7 Cups (peer "listeners") are commonly used [uncertain: not verified: their current age limits, moderation, escalation and data terms].
Steps [synthesis]: (1) Before adding any such service to a plan, find and read its current terms on confidentiality, escalation, age and moderation. (2) Record what each states, with the date checked and the URL, in your directory (4.7). (3) Describe it to the client in the service's own published terms. (4) If its escalation or data terms cannot be found, prefer a service whose terms are published.
Example language (illustrative): "Before we put that one in your plan, let's look up what it currently says about privacy, age and what happens if you tell them you're not safe."
Watch for [synthesis]: relying on a service's reputation instead of its current terms; services aimed at an age group the client is outside. Working: every service in the plan has dated, published terms on file.

4.7 Building and maintaining your service directory

[synthesis, informed by the service pages in 4.1 to 4.5]

Purpose and when to use: before using any service in a plan; refresh at least twice a year and whenever a client reports a change.
Steps: (1) For each jurisdiction your clients live in, list one voice line, one text/chat line, one young people's option and one moderated peer space. (2) For each, record: channels and hours; who responds (volunteers, paid counsellors, clinical supervision); age limits; published confidentiality exceptions; published escalation approach; data retention and research use; how to request deletion; date checked and URL. (3) Add your own service's crisis pathway and emergency numbers. (4) Note channels that have closed (e.g. Samaritans UK email and letters). (5) Bring the directory to supervision once a year.
Example language (illustrative, to a colleague): "I've rechecked the text lines: Shout's privacy notice was updated in June 2026, so I've updated our handout."
Watch for: outdated handouts; services described from memory; US services given to non-US clients. Working: you can answer a client's "what happens if…" question from a dated source.

4.8 Assessing the client's online help-seeking and exposure

[synthesis]

Purpose and when to use: at intake for anyone with suicidal thoughts or self-harm, and whenever risk rises; always with young people.
Steps: (1) Carry out your service's standard suicide and self-harm risk assessment first. (2) Ask which crisis lines, apps, forums, social media accounts and AI chat tools they use when distressed, and when (night-time is common). (3) Ask what they see: recovery-focused content, others' distress, images of self-harm, method talk, encouragement to self-harm or pro-suicide communities, challenges or hoaxes. (4) Ask about effects: do they feel better, worse, urged, competitive, numb? (5) Ask about past crisis-line contacts: helpful? any emergency response? would they use it again? (6) Ask about preferences and barriers: speaking aloud at home, privacy, deafness, language, fear of police or of being sectioned or hospitalised, cost or data worries. (7) Record findings and feed them into the risk formulation.
Example language (illustrative): "Lots of people go online when things get bad, often late at night. Where do you go? … What do you tend to see there, and how do you feel afterwards?"
Watch for: shame or secrecy (normalise first); signs of active encouragement from others, exposure to method information or a pro-suicide community (raises risk; act under your risk procedures and safeguarding duties). Working: the client describes online use frankly and links it to mood.

4.9 Psychoeducation: what a crisis text or chat conversation is like

Purpose and when to use: before the client first uses a text or chat line, to lower the barrier and set accurate expectations.
Steps: (1) Explain the general shape as the service describes it: e.g. for Shout, automated messages first, then a volunteer, about 45 to 60 minutes, aiming at a calm, safe state and a plan [source: Shout How Shout works]; for 988, a counsellor who will ask directly about suicide [source: 988 Suicide Safety Policy]. (2) Explain that there may be a wait, and what to do while waiting (e.g. the next step on their safety plan) [synthesis]. (3) Explain that they can stop at any time (e.g. Crisis Text Line's STOP) [source: Crisis Text Line Privacy Policy]. (4) Explain that it is one conversation, not therapy, and that they can bring it to the next session [synthesis].
Example language (illustrative): "It's a text conversation with a trained person. They'll ask how you're doing and whether you're thinking about suicide; being honest helps them help you. It usually ends with a small plan for the next few hours."
Watch for: expecting a therapist or instant answers; disappointment if the first exchanges are automated. Working: the client can describe what will happen.

4.10 Discussing confidentiality, data and escalation (informed choice)

Purpose and when to use: whenever a service is added to the plan; whenever the client voices fears about police, hospital or records.
Steps: (1) Show the client the service's own current page, or read its stated exceptions together, with the date. (2) Summarise accurately: most services state they keep conversations private except in listed situations such as imminent risk to life, attempts in progress, and child or adult safeguarding [source: 988 Terms; Crisis Text Line, Shout and Samaritans pages]. (3) Explain what they say they try first (safety planning, least invasive options) [source: 988 Suicide Safety Policy; 988 police FAQ; Shout Privacy Notice]. (4) Cover data: records kept, how long, anonymised research or AI-training use, deletion routes [source: Crisis Text Line Privacy Policy; Shout Privacy Notice]. (5) Ask what this means for them, and agree what they would want to happen. (6) Make clear that you cannot guarantee any service's response [synthesis].
Example language (illustrative): "I can't promise you what any service will do. What I can do is show you what they say. Shout says it tries to make a safety plan with you first, and only if that's not possible might a clinician share your details with emergency services. Let's read it together."
Watch for: the client deciding not to use any service because of fear (offer an alternative they will use, and work on the fear); clients choosing to hide risk from responders (discuss the trade-off openly). Working: the client can state the exceptions in their own words and still names a service they would use.

4.11 Adding services to the written safety plan

[synthesis; use the safety-plan format your service uses]

Purpose and when to use: as part of collaborative safety planning for any client with suicidal thoughts or self-harm.
Steps: (1) Place crisis lines at the point in the plan where personal coping and contacting trusted people are not enough, before or alongside emergency services, as the client prefers. (2) Write the exact number or short code and keyword, hours, channel and one line on what to expect. (3) If the client uses a moderated peer space, list it under "things that help me between sessions", not under crisis contacts, with a note on when to step away. (4) Include your service's own crisis pathway and emergency numbers (999/112 in UK, 911 in US). (5) Have the client store the contacts in their phone now, in session. (6) Agree when the plan is reviewed.
Example language (illustrative): "So if the urges are getting stronger and talking to your sister hasn't helped, step four is texting SHOUT to 85258. Let's put it in your phone now, under a name you're comfortable with."
Watch for: plans that list many services but none the client would use; plans that leave out emergency numbers. Working: the client can retrieve the contact quickly and later reports having used the plan.

4.12 Rehearsing the first message

[synthesis]

Purpose and when to use: for clients who freeze when distressed, are new to text lines, or have had a poor first contact.
Steps: (1) Ask the client to draft a short opening text, in their words, that says what is happening and how intense it is. (2) Save it as a note or draft. (3) Role-play the first few exchanges, including being asked about suicide. (4) Practise saying what they need ("I just need to talk" or "I need help staying safe tonight").
Example language (illustrative): "What would you want your first text to say? It can be as short as 'I'm not safe tonight and I don't know what to do.'"
Watch for: perfectionism about wording; drafts that minimise risk. Working: the client sends the message more readily when needed.

4.13 Young people: safeguarding, consent and parent or carer involvement

Purpose and when to use: any client under 18 (and young adults, where helpful).
Steps: (1) Check each service's stated age limits (e.g. Togetherall states 16 and over; Crisis Text Line's policy sets no minimum age or parental consent requirement; the Shout notice states no minimum age) [source: respective pages]. (2) Explain that services publish safeguarding exceptions for children (e.g. Crisis Text Line reports abuse or neglect of under-18s as required by law; Samaritans may contact services if a child is or may be harmed) [source: Crisis Text Line Privacy Policy; Samaritans contact page]. (3) Follow your own jurisdiction's and service's rules on consent, assent, Gillick competence or equivalent, and on sharing with parents or carers [synthesis]. (4) With the young person's agreement and where safe, involve a parent or carer in knowing the plan exists and where help is. (5) Use #chatsafe (developed by Orygen with young people, for ages 12 to 25) for online safety coaching (see 4.15) [source: Orygen 2023 release].
Example language (illustrative): "These text lines are for young people too. Like me, they publish when they would share information, for example if they think you or another young person is being hurt or is in immediate danger. Let's read what this one says. Would it help if your mum knew this number was in your plan?"
Watch for: abuse disclosures (act on your safeguarding duties); a young person keeping use secret because of family conflict. Working: the young person uses the plan and knows who else knows about it.

4.14 Risks of unmoderated forums and pro-self-harm content: a harm-reduction conversation

Purpose and when to use: when assessment shows exposure to self-harm or suicide content, pro-self-harm or pro-suicide communities, or unmoderated forums.
Background: Samaritans' 2020 industry guidance treats self-harm and suicide content online as capable of causing serious harm and calls for platforms to reduce access and increase support [source: Samaritans news release]. In England, a 2022 UK coroner's inquest into a teenager's death led to a Prevention of Future Deaths report (13 October 2022) sent to the government's digital department and to Pinterest, Meta, Snap and Twitter [source: Judiciary UK report page] [uncertain: not verified: the report's findings and recommendations, which are not summarised here].
Steps: [synthesis] (1) Explore without alarm what they see and why they go there (connection, being understood, validation). (2) Name the difference between moderated, recovery-oriented spaces with house rules and escalation (e.g. Togetherall) and spaces that share images, method talk or encouragement. (3) Ask about effects on urges and mood after viewing. (4) Agree concrete steps the client chooses: muting or unfollowing, using platform reporting tools and sensitive-content settings, leaving specific groups, not browsing alone at night, replacing with a moderated space or a crisis line. (5) If the client is being encouraged by others to self-harm or die, or is under 18 and exposed to such communities, treat this as raised risk and a safeguarding matter under your procedures. (6) Never ask for, discuss or repeat method details; focus on effect and avoidance.
Example language (illustrative): "It sounds like that group is the one place you feel understood. I also hear that after you've been on it, the urges are stronger. What would it be like to keep the understanding but lose the part that's making things worse?"
Watch for: bans that increase secrecy; the client acting as a 'helper' to others in crisis; signs of grooming or coercion. Working: reduced exposure, urges after online use ease, the client tells you when they slip back.

4.15 Coaching safer online communication (#chatsafe and safe messaging)

Purpose and when to use: clients, especially young people and content creators, who post or reply about self-harm and suicide.
Source and version: Orygen's #chatsafe guidelines, first edition 2018, second edition May 2023, developed with young people; the second edition adds advice on dangerous online games, challenges and hoaxes and dedicated advice for influencers making mental health content [source: Orygen #chatsafe pages]. Orygen's research lead is quoted that talking about these topics safely "can in fact reduce risk" [source: Orygen 2023 release]. The detailed tips are not summarised here [uncertain: not verified]; give the client the official download and read it together.
Steps: [synthesis, general safe-messaging practice] (1) Read the relevant #chatsafe section together. (2) Before posting: is this the right place, who will see it, am I safe right now? (3) Share feelings and recovery, not methods or images. (4) When someone else posts that they are at risk: respond kindly, encourage them to contact a crisis line or emergency services, use the platform's reporting tool, and tell a trusted adult or professional if they are in danger. (5) Look after yourself: you are not their crisis service.
Example language (illustrative): "There's a guide written with young people about talking online about this stuff. Shall we look at the bit about what to do when a friend posts something scary?"
Watch for: the client carrying responsibility for online friends; posting while acutely distressed. Working: the client describes posting or replying more safely and feeling less burdened.

4.16 Supporting healthy use of moderated peer communities

[synthesis, with Togetherall details from its FAQ]

Purpose and when to use: clients who are isolated, between sessions, or after discharge.
Steps: (1) Check eligibility (often via NHS, employer or university). (2) Discuss anonymity and the advice not to share personal details [source: Togetherall FAQ]. (3) Agree a purpose (e.g. connection, practice of skills, courses). (4) Agree limits: time, not using it as the crisis plan, stepping away from distressing threads. (5) Review in sessions: what helped, what made things worse.
Example language (illustrative): "What would you want to get from it, and how will you know if it's starting to make things harder?"
Watch for: replacement of real-world relationships, comparison, contagion of distress, over-helping. Working: reduced isolation without escalation of urges.

4.17 Debrief and follow-up after a crisis contact or emergency response

[synthesis]

Purpose and when to use: next session after any crisis-line contact, forum crisis or emergency dispatch; sooner if your service offers between-session follow-up.
Steps: (1) Ask what happened, in order. (2) Ask what helped and what didn't. (3) If there was an emergency response, explore the experience, including any distress, shame or loss of trust; validate; check for injuries or new risk. (4) Reassess current risk. (5) Revise the plan (different service, different wording, earlier step). (6) If the client wants to complain or request their data, help them find the service's route (e.g. Samaritans' feedback and complaints page; Crisis Text Line's DELETE or legal contact) [source: Samaritans news release links; Crisis Text Line Privacy Policy].
Example language (illustrative): "You reached out on Saturday night. That took something. Talk me through it: what did they say that helped, and what didn't?"
Watch for: avoidance of all future help-seeking after a distressing response; minimising. Working: the client remains willing to use some crisis pathway.

4.18 Coordinating, documenting and supervision

Purpose and when to use: when you are worried about a client between sessions, need to contact a crisis service yourself, or are recording plans.
Steps: (1) If you contact a crisis line about a client (a third-party contact), know that 988's policy requires counsellors to use the least invasive, most collaborative steps and try to reach the person at risk directly [source: 988 Suicide Safety Policy]. For imminent danger, use emergency services and your own crisis pathway [synthesis]. (2) Document which services are in the plan, the date their terms were discussed, and the client's choices. (3) Take complex cases (refusal of all services, safeguarding, online exploitation) to supervision.
Example language (illustrative, note): "Safety plan updated: client chose Shout 85258 as step 4; privacy and escalation terms discussed using Shout notice (updated June 2026)."
Watch for: sharing client details with services without a basis to do so; relying on a crisis line instead of acting on your own duties. Working: clear records and timely supervision.

5. Handling problems

The services' own pages do not address most of these difficulties; the guidance below is [synthesis] unless a source is given.

"They'll call the police." Do not reassure categorically. Show the service's published approach: e.g. 988 states emergency services are contacted only when risk is imminent or in progress and a less invasive plan cannot be agreed, and lists alternatives; Crisis Text Line states that saying you feel suicidal does not by itself mean emergency services are called [source: 988 police FAQ; Crisis Text Line Privacy Policy]. Then explore the fear's roots (past experiences, racism, immigration status, disability, previous detention), and agree a pathway the client will actually use. Ask: "What would make it possible to reach out, even with that risk?"
Client refuses all services. Accept for now; strengthen personal coping and trusted-person steps; clarify your own crisis pathway; revisit after rapport grows. Document and take to supervision if risk is high.
Bad past experience (long wait, scripted or unhelpful responder, unwanted dispatch). Debrief fully (4.17); validate; consider a different service or channel; rehearse what to ask for (4.12). Offer help with feedback or complaints if wanted.
Over-reliance on crisis lines or forums (many contacts a day, using them in place of sessions or relationships). Explore the function; widen the coping repertoire; agree a pattern of use; involve the service's own guidance where it exists (e.g. Shout offers single conversations) [source: Shout How Shout works].
Client acting as a helper or rescuer to others online. Name the burden; use #chatsafe on responding to others and looking after yourself; agree who the client passes concerns to (platform report, emergency services, a trusted adult) [source: Orygen #chatsafe pages for the guide's existence and scope].
Ongoing exposure to pro-self-harm content or communities, or encouragement by others. Work through 4.14; increase monitoring of risk; for young people or anyone being coerced, follow safeguarding procedures. Avoid shaming or confiscation advice that drives use underground.
Data and privacy distress (e.g. after learning about the Loris.ai episode, or the research and AI-training use of anonymised data). Give dated facts: Crisis Text Line ended the Loris arrangement in January 2022 and asked for the data to be deleted; current policies describe retention, research use and deletion routes [source: Popular Science 2022; Crisis Text Line Privacy Policy; Shout Privacy Notice]. Let the client weigh it; offer a voice line or a different service if preferred.
Wrong country or closed channel. Check location and the directory; do not give 988 or Crisis Text Line to UK clients, or Samaritans email to UK clients now that it has closed [source: Samaritans contact page; Crisis Text Line Privacy Policy].
Minor using adult-only or age-restricted services. Check the service's stated age limit (e.g. Togetherall 16+); redirect to age-appropriate options; involve safeguarding where indicated [source: Togetherall FAQ].
Rupture with you over safety planning ("you just want to hand me off to a hotline"). Acknowledge the feeling; explain that the crisis line is for hours when you are not available, not a replacement for your work together; ask what would make it feel like support rather than dismissal.
Overwhelm in session while discussing online content. Slow down; ground; stop detail-seeking; return to effects and next steps only. Never ask for images or method specifics.
Client uses general-purpose AI chat for emotional support instead of crisis lines. Ask what they use it for and what it offers them; explain that it is not a crisis service or therapy; make sure a human crisis option and emergency numbers are in the plan. See the Domain IX AI chatbot entries for detail.
Clinician uncertainty about a service's current policy. Do not guess: read the current page with the client or say you will check, and record the date.

Sources

Service pages, policies and reports (status checked October 2026):

[uncertain: not verified] Kooth and 7 Cups terms; the full text of the #chatsafe guidelines, Samaritans' technology-industry guidelines and the 2022 coroner's report; Shout's minimum age; Togetherall's emergency-escalation details beyond its FAQ.

← Previous · IX-100Digital phenotyping and passive smartphone sensing Next · IX-102 →Telepsychiatry