AVATAR Therapy Cut Voice Distress at 16 Weeks—Access Remains Limited
AVATAR Therapy now has stronger evidence than the early experiments that first drew attention to the technique. A large randomized trial found that therapist-guided dialogue with a digital representation of a distressing voice reduced voice-related distress at 16 weeks, although the primary benefit was no longer statistically conclusive at 28 weeks.
The treatment has also gained conditional support for adult use in England’s National Health Service while further evidence is collected. That is not the same as routine availability: access remains limited, and the intervention must be delivered within specialist mental-health care rather than used as a self-help program.
What happens during AVATAR Therapy
AVATAR Therapy is intended for people with psychosis who experience persistent, distressing auditory verbal hallucinations. That population includes people with schizophrenia-spectrum diagnoses, but it is broader than schizophrenia alone. Referring to the treatment simply as a conversation between “schizophrenics” and their voices also obscures the clinician’s central role.
The participant and therapist create a digital face and modified voice representing one prominent voice the person hears. In later sessions, the participant sees the avatar on a screen while the therapist alternates between speaking normally and speaking through voice-transformation software. The exchange is therefore a controlled therapeutic dialogue, not an attempt to persuade the participant that the hallucination is an external person.
Early conversations can reproduce the voice’s hostile or controlling manner, allowing the participant to practise answering it more assertively. The therapist gradually changes the avatar’s responses as the participant challenges its authority and develops a stronger sense of control. More personalized sessions may connect the voice’s content and interpersonal power with the participant’s experiences and relationships.
The main target is distress and the person’s relationship with the voice. Improvement can mean feeling safer, less dominated or better able to respond; it can also involve changes in voice frequency or severity. Complete disappearance of the voice is neither guaranteed nor required for the treatment to be useful.
What the AVATAR2 trial found
The peer-reviewed AVATAR2 trial report documents a randomized study of 345 UK adults comparing treatment as usual with treatment as usual plus either a six-session brief protocol or a 12-session extended protocol. Both versions reduced voice-related distress and overall voice severity relative to usual care at 16 weeks, but those between-group differences were not statistically significant at 28 weeks. Only the extended protocol reduced voice frequency at both assessments; completion reached 81.9% in the brief group and 57.9% in the extended group.
The extended course crossed the trial’s predefined threshold for a clinically significant change in distress at 16 weeks. However, the study was not designed to establish whether the extended version was superior to the brief version, so their outcomes should not be treated as a direct head-to-head ranking.
The lower completion rate for the extended course is an important implementation constraint. A more personalized intervention may produce benefits across a wider set of outcomes, but it also asks participants to remain engaged for longer and can include more direct work on difficult life experiences. Future services will need to determine how much therapy can be delivered without losing the people it is meant to help.
No serious adverse event was judged related to the treatment, device or trial procedures. That finding applies to supervised delivery within the study and does not establish that confronting a simulated voice is safe without clinical assessment, monitoring and support.
Why this is not evidence of a cure
AVATAR2 tested the therapy as an addition to usual care, not as a replacement for medication, psychological treatment or crisis support when those are clinically indicated. Its control group received usual treatment rather than cognitive behavioural therapy for psychosis, so the trial cannot show that AVATAR Therapy is better than every established psychological intervention.
The timing of the results also matters. Distress and severity improved within the therapy groups by the later assessment, but the differences from usual care were no longer statistically conclusive. The sustained result for voice frequency was confined to the extended course, and neither protocol produced a significant effect on the trial’s hallucination-remission measure.
The evidence therefore supports a focused, short-term benefit rather than a universal or permanent resolution of psychosis. It does not establish that everyone will respond, that voices will routinely stop or that benefits will continue indefinitely after treatment ends.
Conditional NHS support has not produced routine access
Under NICE’s current recommendations, AVATAR Therapy may be used as an NHS option for adults while more evidence is generated, provided it is delivered or supported by a trained mental-health professional and has the required regulatory and NHS digital-assurance approvals; for young people, access should remain funded through research, company or non-core NHS routes rather than routine NHS use.
This is an early-value recommendation, not approval for automatic nationwide rollout. Outstanding questions include longer-term symptom changes, relapse or deterioration, adverse events, completion, staffing demands, implementation costs and which patients are most likely to benefit. NICE can reconsider routine adoption after reviewing the additional evidence.
The developer’s current access notice states that no centre presently offers direct access to voice hearers, while several NHS centres are planned for 2026–27. Planned openings may change, so conditional national guidance should not be interpreted as confirmation that a local service is already operating.
What has changed since the early studies
The core technique remains recognizable: a private and often dominating experience is represented on screen so that the participant can address it with a therapist’s support. What has changed is the scale of the evidence. The approach has now been tested across multiple UK sites, by a broader clinical workforce and in two protocols with follow-up beyond the treatment period.
The central unresolved issue is no longer whether a digital avatar can produce a measurable therapeutic effect. It is whether ordinary services can reproduce that benefit safely, affordably and equitably—and whether the more demanding extended protocol can retain enough participants to justify its wider range of effects. For now, AVATAR Therapy is best understood as a promising clinician-delivered intervention with conditional policy support, meaningful limitations in the evidence and restricted real-world access.
Also read:
Subscribe to our newsletter
Get the latest Web3, AI, and crypto news delivered straight to your inbox.