MIT researchers develop tool to estimate suicide risk from Text messages
Researchers at MIT’s McGovern Institute, led by former graduate student Daniel Low and senior scientist Satra Ghosh, created a language‑processing tool that estimates suicide risk from text conversations. Using a custom lexicon of about 60 words or phrases for each of 49 risk factors, they analyzed de‑identified data from roughly 16,000 Crisis Text Line chats that were classified into…
Key points
- Researchers analyzed de‑identified texts from ~16,000 Crisis Text Line conversations to train a suicide‑risk prediction model.
- The model uses a custom lexicon of ~60 words/phrases for each of 49 risk factors and outperforms typical symptom cues.
- The lightweight, interpretable model runs on a personal computer; the lexicon and software package are being openly shared.
The team reports that their lightweight machine‑learning model accurately predicts risk severity in new conversations, flagging high‑impact factors such as mentions of lethal means, substance use, active suicidal ideation and self‑injury. Because the model links each term to a specific factor, it remains interpretable and can run on a personal computer, reducing cost and privacy concerns. The researchers are openly sharing both the lexicon and the software package so other scientists can build similar tools for mental‑health conditions, and they see potential for clinical and crisis‑support settings after further validation.
Estimating suicide risk from text
MIT News on AI · 24 September 2026
Loading the full article…
This text was published by MIT News on AI and written by Jennifer Michalowski | McGovern Institute for Brain Research. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- Endura Therapeutics uses AI to cut drug discovery costs by triaging 500 diseases · 1 src
- AI researchers debate tradeoffs of air-gapping models to stop rogue behavior · 1 src
- NVIDIA and Google DeepMind release 2,800 viral protein structures · 1 src
- DR-GEM improves single-cell embeddings and rare cell detection · 1 src
- Researchers unveil UpTCR model for TCR-antigen binding prediction · 1 src
Comments
via GitHub Discussions