Check your content against UK speech law frameworks

Home / Blog / Encouraging Self-Harm Online: What Section 184 Actually Prohibits

← Back to Blog

Encouraging Self-Harm Online: What Section 184 Actually Prohibits

Published 9 July 2026

In 2017, a 14-year-old girl died by suicide. Her family found extensive social media content about self-harm and suicide on her accounts — not just images, but direct encouragement from other users. There was no specific law against it. The platforms weren't legally required to remove it. The individuals who sent the encouragement weren't committing a named criminal offence. They might have been charged under the Communications Act 2003 or the Malicious Communications Act 1988 — but neither statute was designed for this. Neither mentioned self-harm. Neither sent the message that encouraging someone to harm themselves was a distinct, serious crime.

That changed on 31 January 2024, when section 184 of the Online Safety Act 2023 came into force. For the first time, encouraging or assisting the serious self-harm of another person became a specific, named criminal offence — with a maximum sentence of 5 years' imprisonment. This guide explains what the offence covers, what the prosecution must prove, and where the line sits between harmful content and protected speech.

Disclaimer: This is general information, not legal advice. If you or someone you know is affected by self-harm content online, contact the Samaritans (116 123) or PAPYRUS (0800 068 41 41). If you're under investigation under s.184, speak to a solicitor immediately.

Jurisdiction: This covers England and Wales. The offence extends to England, Wales, and Northern Ireland. Scotland has separate equivalent provisions.

Section 184: Encouraging or Assisting Serious Self-Harm 1. COMMUNICATION Sends a message encouraging or assisting self-harm Any electronic communication 2. MENTAL ELEMENT Intent to encourage/assist OR recklessness as to whether Intent or recklessness — not both 3. SERIOUS HARM Self-harm must be "serious" — grievous bodily harm equivalent Not trivial or minor harm No requirement that self-harm actually occurred — the offence is in the sending, not the outcome "Serious" is a legal term of art — it means harm that amounts to grievous bodily harm or worse This does NOT criminalise: discussing self-harm with someone who is struggling, sharing recovery resources, or expressing concern Either-way offence — Maximum 5 years' imprisonment on indictment Sources: Online Safety Act 2023, s.184; Home Office Explanatory Notes
The three elements of the s.184 offence. No actual self-harm needs to have occurred — the offence is complete when the message is sent.

What "Serious" Means — And Why That Word Matters

The most important word in section 184 is "serious." The offence does not criminalise encouraging minor self-harm. It does not criminalise discussing self-harm in general. It does not criminalise telling someone that you understand what they're going through because you've been there yourself. The Act specifically limits the offence to encouragement or assistance of serious self-harm — meaning harm that amounts to, or is equivalent to, grievous bodily harm.

This was a deliberate drafting choice. Parliament was aware that an overly broad offence could criminalise peer support, recovery communities, and even healthcare discussions. The "serious" qualifier is the filter: it separates encouragement that crosses the line into genuine criminality from conversations about self-harm that, while difficult and painful, are not criminal.

In practice, this means a message saying "you should cut yourself" might not meet the threshold if the context suggests the encouragement is toward minor harm. A message saying "here's how to cause injuries that will require hospitalisation" or "you should end your life" almost certainly does. The prosecution must prove that the self-harm encouraged was serious — and that the sender knew it was, or was reckless as to whether it was.

Intent or Recklessness: The Two Routes to Liability

Like s.183 (epilepsy trolling), s.184 allows the prosecution to prove the mental element in one of two ways: intent or recklessness. The prosecution doesn't need to prove the sender actually wanted the recipient to self-harm — it's enough that they were aware of the risk and unreasonably took it.

The intent route is straightforward: the sender meant to encourage serious self-harm. The recklessness route is more nuanced: the sender was aware that their message was likely to encourage serious self-harm, and they sent it anyway — not caring whether it did or not. Neither route requires proving that the recipient actually self-harmed. The offence is about the sender's conduct and state of mind, not the outcome.

There is a statutory defence: if the communication was sent for the purpose of preventing or reducing serious self-harm, it is not an offence. This protects healthcare professionals, helpline volunteers, peer support workers, and anyone who genuinely intends to help. The defence is not a loophole — it's a recognition that the law must distinguish between encouragement and support.

How Section 184 Fits With Other Offences

Section 184 is part of a cluster of harm-based offences in Part 10 of the Online Safety Act 2023. Understanding them together explains why Parliament created this category of crime:

Online Safety Act 2023, s.179 — False Communications: Knowingly false messages intended to cause non-trivial psychological or physical harm. Covers the situation where someone falsely tells a vulnerable person that a loved one has died, intending to cause devastating distress. Maximum: 6 months.

Online Safety Act 2023, s.181 — Threatening Communications: Threats of death or serious harm. Intent or recklessness. Maximum: 5 years.

Online Safety Act 2023, s.183 — Epilepsy Trolling: Sending flashing images with intent to induce a seizure, or recklessness. Maximum: 5 years.

Online Safety Act 2023, s.184 — Encouraging Self-Harm: Encouraging or assisting serious self-harm. Intent or recklessness. Maximum: 5 years.

The common thread: these are not offences about offensive content. They're offences about harm — physical or psychological damage targeted at specific vulnerabilities. Parliament created them alongside the older content-based offences (s.127, Malicious Communications Act) to address conduct that falls between the cracks of the existing law.

Communications Act 2003, s.127: Before s.184, prosecutors could charge encouragement of self-harm under s.127 as "grossly offensive" or "menacing." But it was an imperfect fit — and the maximum was only 6 months. Section 184 gives prosecutors a precise tool with a sentence that reflects the gravity of the conduct.

The Harm-Based Offences: Before and After the OSA 2023 BEFORE OSA 2023 (Pre-Jan 2024) s.127 Comms Act — 6mo max, wrong fit MCA 1988 — purpose to cause distress, 2yr No specific offence for encouraging self-harm → AFTER OSA 2023 (31 Jan 2024 onwards) s.184 OSA — specific offence, 5yr max s.183 — epilepsy trolling, 5yr max Precise tools for specific types of harm WHAT s.184 DOES NOT CRIMINALISE Discussing self-harm with someone who is struggling | Sharing recovery resources | Expressing concern Healthcare discussions | Peer support communities | Journalism and awareness-raising Statutory defence available if the communication was sent to prevent or reduce serious self-harm Sources: Online Safety Act 2023, ss.179, 181, 183, 184
Before and after the OSA 2023: how the harm-based offences filled gaps in the law, with explicit carve-outs for legitimate support.

Practical Takeaways

  1. Encouraging serious self-harm is now a specific crime with a 5-year maximum. The law has moved from relying on older, ill-fitting statutes to a precise, named offence that reflects the gravity of the conduct.
  2. The "serious" qualifier matters. Trivial encouragement of minor harm doesn't meet the threshold. The self-harm encouraged must be equivalent to grievous bodily harm — and the prosecution must prove the sender knew that, or was reckless as to it.
  3. There is a statutory defence for prevention and support. Healthcare professionals, helpline workers, peer supporters, and anyone genuinely trying to prevent self-harm are protected. The law was not designed to criminalise help.
  4. No actual self-harm is required. The offence is complete when the message is sent with the required intent or recklessness. The harm is in the conduct, not the outcome.

Sources

  • Online Safety Act 2023, s.184 — legislation.gov.uk
  • Online Safety Act 2023, ss.179, 181, 183 — legislation.gov.uk
  • CPS Social Media Guidelines — cps.gov.uk
  • Samaritans — 116 123
  • PAPYRUS (young suicide prevention) — 0800 068 41 41

UK Content Comply analyses content against 13 UK speech law frameworks in total. This article focuses on the Online Safety Act 2023 — for coverage of other frameworks such as the Communications Act 2003, the Public Order Act 1986, or the Crime and Policing Act 2026, see our UK Speech Law resources.