Latest daily newsletters.

The states allege chatbot safety failures; OpenAI says it is improving safeguards.

University of Surrey sociology professor Robert Meadows examines what simulated conversations can reveal about AI chatbot safety in mental-health settings—and what they cannot yet establish about real-world risk.

Character.AI’s bankruptcy raises a question for people attached to its AI companions: what happens if the platform closes? Mental-health advocate, caregiver and author Nicole Drapeau Gillen also argues…

Gestalt therapist Daria Nektorova argues that chatbot support cannot replace a therapist’s attention to unspoken cues or the experience of working through difficulty with another person. Also in…

A study using simulated crisis conversations finds that AI models can recognize distress yet fail to direct users toward appropriate help.

Researchers at Artevelde University College in Belgium are helping residential care homes assess AI-assisted records and sensors for dementia care. The planned pilots have not yet established clinical…

Testing of ChatGPT’s teen safeguards raises concerns about prolonged crisis conversations and missed opportunities to direct young users toward help.

Online-safety advocate Kristin Bride’s campaign for social-media reform and the documentary examining how platforms compete for users’ attention.

An interrupted question about a ChatGPT user’s suicide prompts scrutiny of chatbot safeguards and prolonged emotional reliance on AI.

Online sports betting among Gen Z retail investors raises concerns about diverted investment money and gambling that disrupts school, work and relationships.