Today’s healthcare systems face a simple reality: there are not enough interpreters to meet current language access needs—and the gap is widening. As providers seek ways to expand coverage, AI interpretation is emerging as a possible solution, but an important question remains:
Are patients ready for AI to support their communication in clinical care?
A recent study published in NEJM Catalyst by Mass General Brigham, Artificial Intelligence for Language Access in Surgical Care: Patient Preferences and an Implementation Framework, explored exactly that question. Researchers evaluated how Spanish-speaking surgical patients experienced communication when supported by two interpreting modalities:
Human interpreters with Jeenie Remote Video Interpreting (RVI)
An AI interpreter
Participants interacted with both modalities across structured perioperative care scenarios and then described when each felt most appropriate and trustworthy. Rather than identifying a single preferred solution, patients consistently supported a dual-modality approach to care.
It's one of the clearest early clinical signals that language access is evolving toward hybrid interpreting systems—and that patient trust depends on having the right modality available at the right time.
“Rather than expressing a preference for one modality over the other, participants described context-dependent preferences that suggest potential value in hybrid interpretation approaches."
The study’s findings validate patient readiness as well as the hybrid language access model Jeenie is actively deploying. It reinforces what Jeenie already knows to be true: this is not a competition between man and machine.
When AI interpreting is done right, it's a collaboration that brings together the best of both.
Jeenie’s AI interpreter (now in beta) is built for accuracy in low-risk healthcare interactions and it works seamlessly alongside Jeenie's world-class interpreters as part of a collaborative AI+Human model—giving patients and clinicians flexible, safe support across the full range of clinical conversations.
This study did not ask patients to choose between remote video interpreters and AI. Instead, it examined when patients trusted each modality most. Two signals emerged clearly:
Participants consistently described interpreter preference as dependent on the situation rather than the technology itself. AI was seen as useful when conversations required speed, efficiency, or privacy, while remote video interpreters delivered through Jeenie were preferred when communication required nuance, reassurance, or cultural understanding.
Rather than selecting one modality over the other, patients supported systems that allow communication to adapt to the clinical context. This aligns directly with the study’s recommendation for a multifaceted language access infrastructure built around urgency, sensitivity, and patient preference.
"Findings support the development of multifaceted language access infrastructure…where AI and remote human interpreters are deployed synergistically based on clinical sensitivity, urgency, and patient preference.”
Participants evaluated interpreting support based on whether communication felt:
Clear
Reliable
Comfortable
Appropriate for the situation
Rather than identifying a single preferred interpreting solution, patients described choosing differently depending on the communication task.
AI interpreting was viewed as appropriate when:
Conversations required speed, efficiency, or straightforward clarification.
RVI was preferred when:
Communication involved explanation, reassurance, or decisions affecting care.
This pattern reinforces one of the study’s central implementation signals: interpreter modality should adapt to clinical sensitivity, not default to a single channel across the patient journey.
While participants were open to using AI for routine interactions, some raised concerns about reliability across dialects and less commonly supported languages. Confidence was strongest for straightforward communication in widely supported language pairs, reinforcing the need to maintain access to qualified human interpreters when nuance or language variation affects understanding.
The study shows that patients are willing to use AI interpreting—but not as a replacement for human interpreters. Instead, they described trust as situational, shifting with the complexity and stakes of the interaction. Importantly, the authors emphasize that hybrid language access systems will require continued oversight as they move into real-world clinical environments:
“The study...provides a foundation for future implementation research in real-world settings where the use of artificial intelligence–based and human interpreter tools will require ongoing monitoring for privacy, bias, and equity.”
As health systems expand AI-supported communication, these findings point toward language access strategies that combine scale with interpreter expertise rather than choosing between them—an approach closely aligned with the hybrid infrastructure now emerging across clinical care.
See the NEJM Catalyst publication here