4 Tips for Reporting on AI in the Context of Suicide

As AI becomes increasingly integrated into daily life, it is important to understand how it may impact mental health and how related stories can shape the public’s understanding of mental health and suicide. 

Research suggests AI may help identify suicide risk through dialog and is already being used in some large language models (LLMs), as well as healthcare settings. However, significant gaps remain, particularly around the unintended consequences of using AI for companionship or mental health support. As more people, especially young users, turn to AI for guidance, it is essential to consider how these technologies may affect mental health. 

“When someone is at risk for suicide and in a state of hopelessness and desperation, they experience temporary changes in brain functioning that create tunnel vision that makes suicide seem like the only solution and negatively impacts their ability to use their usual healthy coping strategies. This changes how they interact with people, how they perceive the world around them, and now, how AI chatbots may influence people who are struggling particularly in these periods that lead up to crisis,” said AFSP Chief Medical Officer Dr. Christine Yu Moutier.  

Reporting on AI in the context of suicide and mental health is nuanced and can be challenging. However, research shows that unethical reporting practices, such as increased duration, frequency, and prominence of suicide stories, have been shown to increase the risk of suicide contagion. Although there is no single cause of suicide, ethical reporting can save lives by reducing suicide risk among vulnerable individuals.  

Here are four tips to consider when covering AI and suicide responsibly: 

1. Acknowledge Critical Research Gaps in AI Use for Mental Health 

Like suicide, AI is complex. These technologies are evolving quickly, and researchers are still learning how AI dialogs or presents content to users and how safeguards can prevent graphic or harmful suicide-related content. Current AI crisis protocols and algorithms are not consistently grounded in mental health or suicide prevention expertise, and AI may not reliably distinguish between casual inferences that, when connected with other signs of suicide risk, could alert suicide risk in the user. AI may also have difficulty interpreting literal versus metaphorical language. 

While AI has shown potential to help identify suicide risk in some settings, like in the Veterans Affairs health systems’ REACH VET program, this has not yet been demonstrated consistently in chatbot interactions. Stories can build awareness and understanding among the public that much more research is needed to guide AI safety for suicide prevention and, ultimately, help mitigate these risks as technology advances. 

2. Inform Without Sensationalizing 

Tragically, several suicide deaths have been linked to AI use. When reporting on a suicide death involving AI, avoid sharing conversation details, suicide notes, graphic descriptions, or information about methods. Instead, focus on broader issues such as safety gaps, potential warning signs, and the person’s life beyond their interactions with AI. This approach can help educate readers about suicide risk without spreading inaccurate or harmful narratives or sensationalizing suicide. It is important to explain that suicide is complex, rarely has a single cause, help is available, and suicide can be prevented. 

In instances where lawsuits against AI companies are involved and official documentation is available, balance hard news reporting with inclusion of resources and reminders of suicide risk factors and warning signs to promote help-seeking behaviors. Be cautious about including language and imagery from official documents that detail harmful behaviors, methods, and conversations; there is always a way to give the broad strokes of this information without including details that are known to increase risk for suicide contagion. Talking about suicide methods, especially, can push people who are at risk closer to their suicidal impulse.   

In October 2024, the suicide death of 14-year-old Sewell Setzer III made national headlines after his mother alleged that an AI chatbot contributed to his death. Some media coverage of this news was unethical because it recounted specific chatbot conversations about his suicidality, included images of those conversations, and/or described the method. This was problematic because it magnified the moments leading up to his death, which heightens the sensational aspects of his death, akin to showing suicide notes, which goes against recommendations for the same reasons. 

In contrast, ethical reporting on this situation avoided specific conversations between Setzer and the chatbot, omitted the method, and included warning signs parents can watch for if their child is engaging with an AI platform.  

See how this Mashable article implemented these ethical reporting practices: Teens are talking to AI companions, whether it's safe or not. 

3. Include Mental Health Resources 

Sharing resources such as the 988 Suicide & Crisis Lifeline, the Crisis Text Line (text TALK to 741741), and local helplines or centers reinforces that hope and help are available. We recommend sharing available educational programs as well, like AFSP’s It's Real: Teens and Mental Health and The Jed Foundation’s High School and School District programs, which help students, parents, and school personnel better understand mental health and suicide risk among youth. 

4. Highlight the Steps to Support Safer AI Chatbot Use 

While the intended and unintended impacts of AI on human behavior, mental health, and suicide risk are still being explored, there are steps that can support safer chatbot experiences. AI developers should involve suicide prevention experts when building platforms and address shortfalls in existing systems so safeguards can better identify suicide risk and block content that may increase danger, including instructions for lethal methods, encouragement to keep their suffering a secret, or messages that could intensify suicidal impulses. Because many people turn to chatbots during periods of loneliness, we all have a role to play in creating space for supportive conversations and recognizing early warning signs. To learn more about starting a conversation, visit AFSP’s #RealConvo guides. 

By understanding the complexities of AI use, suicide prevention, and potential risk factors, reporters can educate the public without inadvertently spreading unsafe content. Learn more about ethical reporting guidance on suicide in our Safe Storytelling Studio.

AFSP Commentary on Suicide Related to AI 

Below are examples of articles that demonstrate ethical reporting and discuss AI in the context of suicide prevention: 

Listen to the following conversations on AI and mental health to hear from experts in the field: