When British charity The Samaritans was forced to abandon its'Radar' Twitter app in 2014, many in the health community worried that emerging AI technology was poorly suited to the sensitivity of mental illness. The app, designed to read users' tweets for evidence of suicidal thoughts, was criticised for a host of reasons, with one online petition accusing Radar of breaching the privacy of vulnerable Twitter users by alerting everybody – friends and foe alike – of their condition. But three years on, it appears the incident has not halted AI's incursion into psychological healthcare, which artificial intelligence developers believe could be one of their technology's most exciting applications. One such backer is Jim Schwoebel, CEO of US-based NeuroLex, who made headlines last year with his tool to help doctors screen patients for schizophrenia. When Schwoebel's brother developed psychosis, he told The Atlantic last year, doctors required more than 10 primary-care appointments before he was diagnosed.

People convey meaning by what they say as well as how they say it: Tone, word choice and the length of a phrase are all crucial cues to understanding what's going on in someone's mind. When a psychiatrist or psychologist examines a person, they listen for these signals to get a sense of their wellbeing, drawing on past experience to guide their judgment. Researchers are now applying that same approach, with the help of machine learning, to diagnose people with mental disorders. In 2015, a team of researchers developed an AI model that correctly predicted which members of a group of young people would develop psychosis--a major feature of schizophrenia--by analyzing transcripts of their speech. This model focused on tell-tale verbal tics of psychosis: short sentences, confusing, frequent use of words like "this," "that," and "a," as well as a muddled sense of meaning from one sentence to the next.

People convey meaning by what they say as well as how they say it: Tone, word choice, and the length of a phrase are all crucial cues to understanding what's going on in someone's mind. When a psychiatrist or psychologist examines a person, they listen for these signals to get a sense of their wellbeing, drawing on past experience to guide their judgment. Researchers are now applying that same approach, with the help of machine learning, to diagnose people with mental disorders. In 2015, a team of researchers developed an AI model that correctly predicted which members of a group of young people would develop psychosis--a major feature of schizophrenia--by analyzing transcripts of their speech. This model focused on tell-tale verbal tics of psychosis: short sentences, confusing, frequent use of words like "this," "that," and "a," as well as a muddled sense of meaning from one sentence to the next.