Samaritans

Samaritans work to reduce the risk of suicide and support people in moments of crisis. As more conversations move online, digital spaces, including those shaped by AI and chatbot technologies, play an increasingly important role in how people experience distress, seek help, and encounter harm.

Online language, imagery, and behaviours are evolving rapidly, and emerging technologies can unintentionally expose people to harmful or triggering content. To continue influencing effective policy, guidance, and platform practice, Samaritans needed deeper, evidence-based insight into how suicide and self-harm related narratives appear and spread online. This includes an understanding into where current safeguards may be falling short.

Samaritans worked with Nisien.ai to explore online safety trends using a research-led approach that combines AI, data science, and criminology.

The analysis examined public online discourse across multiple platforms, with a particular focus on:

• Emerging patterns of suicide and self-harm related content
• The growing role of AI systems and chatbots in online interactions
• How language, imagery, and platform design can influence vulnerability and risk
  

The work went beyond volume metrics, helping Samaritans to understand context, narrative evolution, and opportunities for earlier, more effective intervention, particularly in AI-mediated environments.

The insights generated are now being used to:

• Support Samaritans’ internal review of online safety work, reinforcing the need for continued focus as digital risks increase rather than diminish.
• Inform Samaritans’ engagement with Ofcom and contribute evidence to the implementation of the Online Safety Act, especially around emerging risks linked to AI and chatbots.
• Update their industry guidelines, ensuring they reflect the realities of a rapidly evolving digital landscape.
• Engage directly with platforms, sharing evidence that helps strengthen suicide and self-harm prevention policies, remove harmful content, and improve user safety.