Scalable, low-cost interventions promise to reach global populations that traditional services cannot, particularly during humanitarian crises, displacement, and conflict. Yet most digital mental health tools are conceived, designed, and tested in relatively stable Western contexts before being deployed elsewhere [1-3].

This development pattern introduces cultural distance at precisely the point where sensitivity matters most. Evidence of clinical efficacy does not ensure cultural legitimacy, trust, or safety [4]. When interventions cross geopolitical and cultural boundaries, they enter environments shaped by historical trauma, media narratives, power asymmetries, and collective memory. These contextual forces rarely appear in design specifications, yet they profoundly shape how digital tools are perceived and used.

The challenge, then, is achieving scale without reproducing blind spots that undermine efficacy.

Most cross-cultural digital health efforts rely on cultural adaptation frameworks [5]. These typically involve translating content, substituting examples, and adjusting tone to fit a new population. While necessary, these steps assume that meaning is largely transferable and remains intact once linguistic barriers are removed.

Our experience suggests otherwise. When conducted without lived experience embedded in the design process, cultural adaptation risks addressing surface differences while missing deeper layers of meaning.

In contrast, lived-experience co-development treats those with direct experience not as informants, but as co-designers. This approach requires cultural and epistemic humility [6]: acknowledgment that certain insights cannot be inferred, researched, or validated externally. They must be shared by those who live within the context the intervention seeks to address.

The collaboration at the center of this article emerged when the Ukrainian team—displaced by war, living in Ukraine, or actively deployed—approached our group at Evolution Health to help build a digital mental health resource for Ukrainians living in displacement and uncertainty.

These developers were active contributors throughout development, shaping decisions about language, tone, pacing, and presentation. Their technical expertise allowed them to participate fully in design discussions, while their lived experience grounded those discussions in their reality. Decisions were no longer guided solely by evidence hierarchies or best practices, but by continuous dialog about how content would be interpreted emotionally, culturally, and symbolically by people living with war-related disruption.

The divergences were striking. Even when words appeared similar, their emotional weight and associations differed substantially. Terms that Western designers associated with isolation, anger, exhaustion, disconnectedness, or optimism carried layered meanings for the Ukrainian team. Some evoked loss. Others echoed bureaucratic language encountered during displacement. Still others were associated with political messaging or wartime media narratives.

Two examples help illustrate the limitations of conventional adaptation and the benefits of co-design.

The first involved a simple vocabulary exercise. The North American team generated a word cloud, and the Ukrainian team was asked to identify words that felt emotionally relevant, neutral, or problematic within the context of displacement and war and to add or remove words accordingly.

The second example involved an introductory video developed using an artificial intelligence (AI)–generated Ukrainian-language voice-over. From a Western perspective, the voice sounded professional, neutral, and supportive—a practical solution that aligned with common digital production practices.

The video was swapped, however, when the Ukrainian team explained that it was likely to evoke distrust. Synthetic Ukrainian voices had become widely associated with Russian disinformation campaigns, where AI-generated Ukrainian speech was routinely used to deliver propaganda.

This risk was invisible to the non-Ukrainian team. Traditional usability testing might have detected disengagement, but only after trust had already been compromised. The issue was not usability, but cultural safety.

These examples underscore two central lessons: that language is not merely semantic, and that trust is built on signals of authenticity, safety, and legitimacy that are culturally situated. Without lived experience in design, even carefully translated content and well-intentioned choices can misfire and risk harm.

Co-development reshaped the intervention in several ways.

First, meaning consistently took precedence over literal accuracy. Language that was technically correct but emotionally misaligned was revised or discarded. The goal shifted from fidelity to original phrasing toward fidelity to lived interpretation.

Second, engagement strategies were reconsidered. Structures intended to support participation occasionally felt directive or intrusive within a displacement context [7]. Lived experience helped distinguish between supportive guidance and unwanted pressure, particularly for individuals navigating chronic uncertainty and loss of agency.

Finally, harm prevention became an explicit design outcome. Lived-experience contributors identified risks that would not have appeared in formal risk assessments—symbolic associations, emotional triggers, and trust signals embedded in design choices. These insights allowed potential harms to be addressed before deployment, rather than retroactively.

There are ethical responsibilities inherent in working with individuals who are actively living through the experiences an intervention seeks to address. Co-development can be experienced as empowering and meaningful, but it also raises questions about psychological safety that appear insufficiently explored in the digital mental health literature [7].

Even when voluntary, participation can involve exposure to emotionally charged language and experiences, increasing the potential for distress. At present, there is limited empirical guidance on how to assess, monitor, or mitigate potential psychological risks for lived-experience contributors in digital intervention development [8].

This collaboration revealed limits in what even experienced digital mental health designers can know from the outside. Lived experience did not simply improve the intervention; it fundamentally reshaped it. It prevented harm, deepened resonance, and challenged assumptions embedded in the original, North American content.

As digital mental health tools continue to scale, lived-experience co-development should be treated as a methodological and ethical necessity rather than an optional enhancement [9]. Even technically sound, evidence-based design alone cannot prevent contextual misalignment and ensure cultural legitimacy, trust, or safety.

Researchers and designers must consider how lived experience is integrated, compensated, sustained, and supported throughout development. Funders and policymakers should evaluate interventions not only on efficacy, but on how they address cultural meaning and safety.

Ultimately, if digital mental health interventions are to be trusted, culturally legitimate, and safe at global scale, the field must move beyond designing for others toward designing with them.