Artificial Intelligence in Addressing Interpersonal Violence Against Vulnerable Populations: A Scoping Review.
Hallessa de Fátima da Silva Pimentel, Julia Sunmy Aguni Chunga, Ana Maria José Garcia João Pascoal and 3 others
PMID 42659509WHAT IT FOUND
Some studies reported AI tools with 75% to 97% success in identifying or predicting violence cases.
Trauma, privacy, and bias risks remained. Clinicians should not assume these tools are proven safe or effective.
Key findings
01The review included ten studies of AI applications for interpersonal violence against vulnerable populations in healthcare services.
02Included studies reported machine learning and natural language processing models with success rates between 75% and 97% for predicting or identifying interpersonal violence.
03The review found that AI was described as an ally rather than a replacement for human analysis, while automated interactions could reactivate trauma.
STILL TO COME
How it was doneWhat they foundWhat it means for RNs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The review included only ten studies, and many were theoretical essays, reviews, methodological papers, or assessments rather than patient outcome trials. The included studies used very different designs, populations, and AI models, so the review could not compare their results directly or determine which measures were successful. The authors searched main health databases and grey literature, but relevant studies may have been indexed elsewhere and missed. The review says little about some vulnerable groups named in its aim, including black people, LGBTQIA+ people, and people experiencing homelessness, because few included studies addressed them. Most included studies were published in English, which may limit generalizability.
Declared interests
The paper reports funding from the Brazilian Ministry of Health/DECIT/CNPq in partnership with the Gates Foundation. No other conflict-of-interest statement is provided.
The easy way to misread this
Do not conclude that AI tools are proven to reduce violence or improve care. The review mapped heterogeneous studies, including theoretical essays and model assessments, and did not test patient outcomes or show that these tools are safe for routine use.