Should digital health trials be placebo controlled?

Digital health interventions—including mobile apps, software-based therapies and artificial intelligence (AI) chatbots—are proliferating rapidly in clinical care. However, the pace of innovation has outstripped the development of rigorous evaluation methods, particularly when it comes to designing appropriate control conditions. Like other non-pharmacological interventions (eg, physiotherapy, psychotherapy or surgery), digital tools pose significant methodological challenges for placebo design. But unlike those fields, digital health lacks widely accepted standards for constructing ethical and scientifically credible control groups. This gap presents not only a threat to internal validity but also an ethical failure. Inadequate controls can inflate perceived efficacy, obscure harms and contribute to injustice—particularly among underserved populations who may be more likely to use digital solutions. We argue that rigorous control design is not merely a technical necessity but an ethical imperative. In this paper, we review methodological challenges inherited from non-pharmacological clinical trial research, explore the unique complexities of trial design in digital interventions and provide both methodological and ethical recommendations for enhancing the conceptualisation and application of placebo controls in the rapidly evolving field of AI-enabled tools.