DAFNI Project: an AI to Deal with Reporting on Suicide in the Media

Jerusalem, 23 July, 2025 (TPS-IL) — The DAFNI system was developed with the understanding that Media coverage of suicide has a dramatic impact on the public. Studies show that careless coverage, which glorifies suicide or reveals dramatic details, can lead to an increase in suicides.

Recent studies have consistently shown that careless coverage that glorifies suicide or reveals overly dramatic details can lead to an increase in suicides, a phenomenon known as the “Werther effect.” Conversely, sensitive coverage that includes messages of hope and recovery can reduce suicidal ideation and encourage help-seeking—the so-called “Pappegano effect.”

Despite clear guidelines from international bodies such as the World Health Organization (WHO), the guidelines are only partially implemented. The reasons for this include a lack of training for journalists, high turnover in new systems, a lack of real-time professional oversight, and difficulties in systematically assessing the quality of coverage.

To address these challenges, Professor Yossi Levy-Blaz and Professor Zohar Eliosaf, from the Lior Tzafati Center for the Study of Suicidality and Mental Pain at the Ruppin Academic Center and the University of Haifa, developed DAFNI – an advanced artificial intelligence system (GenAI), designed to provide journalists and editors with automatic assessment and correction tools for articles dealing with suicide. The system allows for real-time identification of deviations from guidelines, assessment of the level of compliance with professional standards, and suggestion of corrections to improve the text. The corrections include, among other things, reducing dramatic or romantic wording, omitting descriptions of the method and place of suicide, adding information to help readers, and balancing the description of the story with a preventive and empowering message.