Large language models (LLMs) typically generate identical or similar responses for all users given the same prompt, posing serious safety risks in high-stakes applications where user vulnerabilities differ widely. Existing safety evaluations primarily rely on context-independent metrics--such as factuality, bias, or toxicity--overlooking the fact that the same response may carry divergent risks depending on the user's background or condition. We introduce "personalized safety" to fill this gap and present PENGUIN--a benchmark comprising 14,000scenarios across seven sensitive domains with both context-rich and context-free variants. Evaluating six leading LLMs, we demonstrate that personalized user information significantly improves safety scores by 43.2%, confirming the effectiveness of personalization in safety alignment. However, not all context attributes contribute equally to safety enhancement. To address this, we develop RAISE--a training-free, two-stage agent framework that strategically acquires user-specific background. RAISE improves safety scores by up to 31.6%over six vanilla LLMs, while maintaining a low interaction cost of just 2.7 user queries on average. Our findings highlight the importance of selective information gathering in safety-critical domains and offer a practical solution for personalizing LLM responses without model retraining. This work establishes a foundation for safety research that adapts to individual user contexts rather than assuming a universal harm standard.

Author Note Correspondence concerning this article should be addressed to Yuchen Cao, Northeastern University, E-mail: cao.yuch@northeastern.edu Abstract Social media has become an important source for understanding mental health, providing researchers a way to detect conditions like depression from user-generated posts. This tutorial provides practical guidance to address common challenges in applying machine learning and deep learning methods for mental health detection on these platforms. It focuses on strategies for working with diverse datasets, improving text preprocessing, and addressing issues such as imbalanced data and model evaluation. Real-world examples and step-by-step instructions demonstrate how to apply these techniques effectively, with an emphasis on transparency, reproducibility, and ethical considerations. By sharing these approaches, this tutorial aims to help researchers build more reliable and widely applicable models for mental health research, contributing to better tools for early detection and intervention. Tutorial on Using Machine Learning and Deep Learning Models for Mental Illness Detection Introduction Mental health disorders, especially depression, have become a significant concern worldwide, affecting millions of individuals across diverse populations (Organization, 2020). Early detection of depression is crucial, as it can lead to timely treatment and better long-term outcomes. In today's digital age, social media platforms such as X(Twitter), Facebook, and Reddit provide a unique opportunity to study mental health. People often share their thoughts and emotions on these platforms, making them a valuable source for understanding mental health patterns (De Choudhury et al., 2013; Guntuku et al., 2017). Recent advances in computational methods, particularly machine learning (ML) and deep learning (DL), have shown promise in analyzing social media data to detect signs of depression. These techniques can uncover patterns in language use, emotions, and behaviors that may indicate mental health challenges (Shatte et al., 2020; Yazdavar et al., 2020).

The timely identification of mental disorders in adolescents is a global public health challenge.Single factor is difficult to detect the abnormality due to its complex and subtle nature. Additionally, the generalized multimodal Computer-Aided Screening (CAS) systems with interactive robots for adolescent mental disorders are not available. Here, we design an android application with mini-games and chat recording deployed in a portable robot to screen 3,783 middle school students and construct the multimodal screening dataset, including facial images, physiological signs, voice recordings, and textual transcripts.We develop a model called GAME (Generalized Model with Attention and Multimodal EmbraceNet) with novel attention mechanism that integrates cross-modal features into the model. GAME evaluates adolescent mental conditions with high accuracy (73.34%-92.77%) and F1-Score (71.32%-91.06%).We find each modality contributes dynamically to the mental disorders screening and comorbidities among various mental disorders, indicating the feasibility of explainable model. This study provides a system capable of acquiring multimodal information and constructs a generalized multimodal integration algorithm with novel attention mechanisms for the early screening of adolescent mental disorders.

Scientists have revealed they're planning to create a smartphone app that detects if someone's depressed based on changes in their voice. Speech coordination changes when a person becomes depressed, according to the researchers, at the University of Maryland. Depressed people cannot think as fast, and their speaking rate is slowed with more and longer pauses than if they are not depressed, they say. Therefore, a voice detection app using deep learning – a type of machine learning based on artificial neural networks – could help detect such traits, which can often be subtle. The app could be recommended by mental health therapists to their patients, who would submit video and audio updates on their mood at home, which the technology would then assess.

Digital phenotyping approaches that collect and analyze Smartphone-user data on locations, activities, and even feelings - combined with machine learning to recognize patterns and make predictions from the data - have emerged as promising tools for monitoring patients with psychosis spectrum illnesses, according to a report in the September/October issue of Harvard Review of Psychiatry. The journal is published in the Lippincott portfolio by Wolters Kluwer.John Tourous, MD, MBI, of Harvard Medical School and colleagues reviewed available evidence on digital phenotyping and machine learning to improve care for people living with schizophrenia, bipolar disorder, and related illnesses. Digital phenotyping provides a much-needed bridge between patients' symptomatology and the behaviors that can be used to assess and monitor psychiatric disorders." "Digital phenotyping is the use of data from smartphones and wearables collected in situ for capturing a digital expression of human behaviors," according to the authors. Psychiatry researchers think that collecting and analyzing this kind of behavioral information might be useful in understanding how patients with severe mental illness are functioning in everyday life outside of the clinic or lab - in particular, to assess symptoms and predict clinical relapses.

The Centre for Data Analytics and Cognition (CDAC) at Australia's La Trobe University worked with international cancer researchers to develop an artificial intelligence patient-reported information multidimensional framework to help detect and analyze a patient's mental health status while undergoing cancer treatment. The Centre for Data Analytics and Cognition (CDAC) at Australia's La Trobe University has teamed up with international cancer researchers to develop an artificial intelligence patient-reported information multidimensional framework (PRIME) to detect and analyze a patient's mental health status while undergoing cancer treatment. According to CDAC director and La Trobe University head of analytics discipline, Damminda Alahakoon, using PRIME can help understand a patient's behaviour, emotions, and decision-making based on data shared by the patient. He said the data can be text provided by a patient to an online chatbot, an online cancer support group, or other online support services. "PRIME addresses the challenges associated with understanding the unlabelled and unstructured nature of this data, allowing it to efficiently identify trends and anomalies -- such as when a patient is struggling emotionally -- and effectively adapt to the changing nature of that data," he said.

Thanks to advances in artificial intelligence, computers can now assist doctors in diagnosing disease and help monitor patient sleep patterns and vital signs from hundreds of miles away. Now, CU Boulder researchers are working to apply machine learning to psychiatry, with a speech-based mobile app that can categorize a patient's mental health status as well as or better than a human can. "We are not in any way trying to replace clinicians," says Peter Foltz, a research professor at the Institute of Cognitive Science and co-author of a new paper in Schizophrenia Bulletin that lays out the promise and potential pitfalls of AI in psychiatry. "But we do believe we can create tools that will allow them to better monitor their patients." Nearly one in five U.S. adults lives with a mental illness, many in remote areas where access to psychiatrists or psychologists is scarce.

Like footprints in the snow, a distinct trail remains every time someone accesses a digital device that follows them from website to website and from app to app. While it would seem time spent online--for fun or work--is a person's own, it turns out these digital footprints do not serve them as much as they serve the artificial intelligence constantly looking over their shoulders and gathering that information for later use. Each time someone taps on a smartphone screen--which the average person does more than 2,600 times every day--the data accessed is tracked and stored. Much of the information left behind on the internet is widely available to anyone with the interest and (moderate) computer skills to track it down. Social media sites, where many people often share very personal details and input detailed user profiles, are used to extract information which is later used to target ads.

Subjective Well-being(SWB), which refers to how people experience the quality of their lives, is of great use to public policy-makers as well as economic, sociological research, etc. Traditionally, the measurement of SWB relies on time-consuming and costly self-report questionnaires. Nowadays, people are motivated to share their experiences and feelings on social media, so we propose to sense SWB from the vast user generated data on social media. By utilizing 1785 users' social media data with SWB labels, we train machine learning models that are able to "sense" individual SWB from users' social media. Our model, which attains the state-by-art prediction accuracy, can then be used to identify SWB of large population of social media users in time with very low cost.