If you're evaluating these modalities from an investment perspective, the key distinction is between clinical evidence, engagement, and commercial momentum. Those don't always align.
Here's where the evidence stands today:
| Modality | Evidence strength | Best-supported use cases | Investment outlook |
|---|
| Digital CBT | Strongest | Mild-to-moderate depression, anxiety, insomnia | Mature but still attractive in specialized niches |
| AI coaching/chatbots | Moderate, emerging | Psychoeducation, CBT exercises, symptom monitoring, between-session support | High growth potential, but clinical validation still catching up |
| Virtual reality | Moderate for specific disorders | Phobias, PTSD exposure therapy, social anxiety, pain | Promising but narrower market and implementation challenges |
1. Digital CBT: Still the strongest evidence base
Digital cognitive behavioral therapy (dCBT) remains the most evidence-backed digital mental health intervention.
Multiple randomized controlled trials and meta-analyses consistently show meaningful improvements for:
- depression
- generalized anxiety
- panic disorder
- insomnia (particularly strong evidence)
- relapse prevention
The strongest products closely mirror evidence-based CBT rather than offering general wellness content. Outcomes improve further when digital CBT includes some degree of clinician support or coaching.
Why investors still like it:
- predictable regulatory pathway
- measurable clinical endpoints
- reimbursement opportunities
- relatively low technical risk
The downside is that this market is becoming crowded, making differentiation harder unless companies focus on particular populations or integrate with healthcare systems. Recent reviews continue to identify digital CBT as the best-validated digital therapeutic approach.
2. AI coaching: Rapid progress, but evidence is newer
Generative AI has changed the landscape dramatically.
Early studies suggest AI mental health chatbots can:
- reduce mild anxiety and depression symptoms
- improve engagement
- increase access through 24/7 availability
- encourage completion of CBT exercises
Meta-analyses now show statistically significant—but generally modest—clinical benefits. However, compared with digital CBT, the evidence base is:
- much newer
- more heterogeneous
- based on fewer high-quality RCTs
- lacking long-term follow-up
Important caveats include:
- inconsistent handling of higher-risk situations
- limited evidence for severe mental illness
- regulatory uncertainty
- ongoing concerns about hallucinations, inappropriate reassurance, and overdependence
Many experts increasingly view AI as a clinical copilot rather than a replacement therapist—augmenting homework, symptom tracking, journaling, and between-session support rather than delivering independent psychotherapy.
3. Virtual reality: Excellent evidence—but for targeted indications
VR has one of the more interesting evidence profiles.
For conditions requiring exposure therapy, evidence is surprisingly robust.
Strongest support:
- specific phobias
- fear of flying
- public speaking anxiety
- social anxiety
- some PTSD applications
The reason is straightforward: VR recreates controlled exposure scenarios that are difficult or expensive to reproduce in real life.
Evidence is weaker or more mixed for:
- major depression
- generalized anxiety
- schizophrenia
- eating disorders (promising but still experimental)
The largest implementation barriers remain:
- hardware costs
- clinician training
- workflow integration
- reimbursement
Systematic reviews conclude that VR works well when it directly reproduces established exposure-based therapies, rather than trying to invent entirely new treatment models.
Where investors appear to be shifting
The most attractive opportunities increasingly combine modalities rather than betting on one alone.
Examples include:
- Digital CBT + AI coach for daily adherence
- Human therapist + AI between sessions
- VR exposure + clinician supervision
- Passive monitoring + personalized CBT interventions
The technology itself is becoming less of a differentiator than how effectively it integrates into evidence-based care pathways.
If I were ranking them by current evidence
- Digital CBT — strongest clinical evidence across multiple common mental health conditions.
- VR for exposure-based therapy — strong evidence within a narrower set of disorders, especially phobias and some anxiety conditions.
- AI coaching/chatbots — encouraging early evidence and strong commercial momentum, but still building the long-term clinical evidence base and safety record.
From an investment perspective, many analysts now view the most durable businesses as those using AI to enhance proven therapies (like CBT or clinician-led care) rather than positioning AI as a standalone replacement for therapy. That approach aligns better with the current evidence, regulatory direction, and clinical practice.
