Samaritans set a reference standard for covering suicide safely

The design

Samaritans, the UK suicide-prevention charity, publishes a set of media guidelines on how to report on suicide and self-harm without increasing risk to readers. According to its media guidelines hub, as retrieved on 16 September 2026, the guidance spans more than a dozen specific situations, including inquests, novel methods, celebrity deaths, broadcast, drama and documentary, plus a dedicated page on reporting on self-harm and suicide content online. That page states that exposure to graphic online content “can be distressing and may encourage or exacerbate these behaviours,” and recommends signposting audiences to helpline numbers such as Samaritans' own 116 123 line rather than leaving a report without a route to support.

What the evidence says

Both pages are Samaritans' own published guidance, written primarily to advise journalists and other media producers, not chatbot developers. Neither page mentions AI companions, chatbots, or any named companion product; no claim is made here that any AI company has formally adopted this guidance, because none of the pages checked for this entry states that. Where a chatbot company describes consulting outside expertise on crisis response, as some entries elsewhere in this archive record, this document is offered as a named reference standard such a team could be drawing on, not evidence that any particular one is.

What it asks of people

The guidance asks anyone covering this subject to avoid describing a method in detail, to avoid sensationalizing a death, and to include information about where a struggling reader can get help. Samaritans also makes a media advice team available by email to answer questions from journalists working on a story, extending the guidance beyond a static document into an ongoing point of contact.

Privacy and safeguards

The guidelines are voluntary and carry no enforcement mechanism of their own; Samaritans is a charity, not a regulator, and nothing on the pages checked here claims legal force. What the guidance protects is a reader's exposure to distressing detail and their access to a next step; what it cannot do on its own is verify that any particular chatbot, or any journalist, actually follows it.

• Which, if any, AI companion companies have stated publicly that they consult Samaritans' guidance specifically?
• Does advice written for one-off news articles translate cleanly to an ongoing, personalized chat conversation?
• Who checks whether a chatbot's crisis response actually matches the guidance's recommendations?

Samaritans' guidance is a considered standard built for a different medium than a chat interface; treating it as a template for chatbot design, absent a company's own statement to that effect, is an editorial extension this entry does not make.