Background. Evidence for the efficacy of computer-based psychological interventions is growing. A number of such interventions have been found to be effective, especially for mild to moderate cases. They largely rely on psychoeducation and ‘homework tasks’, and are specific to certain diagnoses (e.g. depression). Aims. This paper presents the results of a web-based randomised controlled trial of Manage Your Life Online (MYLO), a program that
uses artificial intelligence to engage the participant in a conversation across any problem topic. Method. Healthy volunteers (n = 213) completed a baseline questionnaire and were randomised to the MYLO program or to an active control condition where they used the program ELIZA, which emulates a Rogerian psychotherapist. Participants completed a single session before completing post-study and two-week follow-up measures. Results. Analyses were per protocol with intent to follow-up. Both programs were associated with improvements in problem distress, anxiety and depression post intervention, and again two weeks later, but MYLO was not found to be more effective than ELIZA. MYLO was rated as significantly more helpful than ELIZA, but there was no main effect of intervention on problem resolution. Conclusions. Findings are consistent with those of a previous smaller, lab-based trial and provide support for the acceptability and effectiveness of MYLO delivered over the internet for a non-clinical sample. The lack of a no-treatment control condition means that the effect of spontaneous recovery cannot be ruled out.