AI Chatbots Often Fail To Help Users In Mental-Health Crisis
New research shared exclusively with TIME shows chatbots are good at spotting distress. They frequently fail to refer users to resources.
Read the original →Every story in this issue with a short summary, so you can decide what to read before leaving the site. Summaries are written by AI (Claude) only from each publisher's own description and checked by Jev for anything not in that description. They are not full translations. Follow "Read the original" for the full story.

New research shared exclusively with TIME shows chatbots are good at spotting distress. They frequently fail to refer users to resources.
Read the original →This chart shows how users and non-users of AI for emotional support think companies should protect users from potential harm.
Read the original →A scoping review of 19 studies finds that multimodal AI fusing raw images with non-image data in mental and cognitive healthcare is dominated by a single dataset and intermediate fusion strategies.
Read the original →One in three young Canadians say they regularly use generative AI tools at least once a month to support their mental health. A new survey said so on Friday.
Read the original →The commentary extends Tannús and colleagues' AI-driven rehabilitation exergame with a closed-loop adaptive framework that integrates real-time heart rate monitoring for cardiovascular safety. It outlines a three-phase validation roadmap. It describes a pathway to turn passive motor assessment into responsive, safe home rehabilitation for stroke survivors.
Read the original →Meta is donating 1,000 Ray-Ban Meta AI glasses to disability organisations in Singapore. The donation comes with funding for training and practical accessibility support.
Read the original →Jio plans AI wearables with Indian language support, and Torchit is developing smart glasses for accessibility. Qualcomm sees payments and education as key opportunities.
Read the original →The Federal Government has unveiled a 3D-printed prosthetics technology initiative. It aims to modernise rehabilitation services, strengthen local technical capacity and reduce Nigeria's reliance on imported assistive devices.
Read the original →Researchers in China have built a deep-learning model that fuses CT, MRI, and 3D radiation dose data. The model predicts which nasopharyngeal carcinoma patients will develop life-threatening tissue necrosis.
Read the original →AI-powered remote cardiac monitoring helps healthcare providers manage cardiac data, reduce alert overload, predict deterioration, and improve care delivery.
Read the original →NewDays, a Seattle startup using a generative AI therapy to treat people with mild dementia, has raised additional funding and closed its seed round with $16 million. The funding was led by Madrona and General Catalyst. The company also announced on Wednesday that it has expanded its services to Nevada, joining Washington, California, Florida and Texas.
Read the original →An unsupervised machine learning analysis of 23 physical performance measures in 368 older adults identified three performance clusters. The clusters aligned with frailty, fall risk, and living situation.
Read the original →MVision body analyzes existing CT images with AI to visualize fat mass, muscle mass and muscle quality. The press release from Company M (Kabushiki Kaisha M) is dated October 9, 2026, 16:50.
Read the original →