The landscape of mental healthcare is shifting beneath our feet. What once seemed like a distant future, where artificial intelligence collaborates with clinical judgment and digital therapeutics sit alongside traditional therapy on the treatment menu, is now firmly rooted in the present. We're not simply witnessing incremental change; we're watching structural transformation unfold in real time.

If you're a practitioner, clinical director, or clinic owner operating in 2026, you've likely felt the pressure to adapt. New reimbursement pathways are opening. Clients are arriving with smartwatch data and AI-generated mental health recommendations. Organisations that once viewed digital tools as supplementary are now building them into their core operating model. The question is no longer whether digital mental health matters. It's which trends matter most, and how you should respond.

We've examined the evidence, reviewed recent regulatory shifts, and spoken with leaders reshaping the sector. Here are five trends that genuinely stand to reshape your practice in 2026 and beyond.

1. AI-Driven Personalisation Is Becoming Clinically Feasible (But Not Without Questions)

Generative AI in mental healthcare remains contentious, and rightly so. In November 2025, the FDA's Digital Health Advisory Committee convened to discuss artificial intelligence in mental health devices. Regulators flagged legitimate concerns: algorithmic bias, hallucination risks, and what researchers call "sycophancy," where AI systems tell users what they want to hear rather than what they need to hear.

Yet here's what's striking: despite these red flags, nearly 45% of new digital mental health studies published in 2024 incorporated LLM-based chatbots. Only 16% of these included clinical validation. The field is racing ahead of the evidence base.

Where clinical evidence does exist, results are encouraging. Woebot, an AI-supported conversational agent, demonstrated an effect size of d=0.44 for depression symptoms in clients. Wysa achieved d=0.47. But the headline win came in late 2025 when Therabot became the subject of the first randomised controlled trial of a generative AI chatbot in a peer-reviewed medical journal. All participants who completed the 8-week intervention showed marked symptom reduction, a finding the New England Journal of Medicine AI flagged as particularly notable.

The momentum is undeniable. Recent coverage in the APA Monitor (2026) highlighted emerging research where AI systems predict depression risk by analysing physiological signals, activity patterns, and sleep data collected from wearable devices. The ability to detect deterioration before symptoms become obvious shifts the entire value proposition.

What this means for practitioners: AI isn't replacing therapeutic judgment. It's expanding capacity and accelerating insight. Digital tools can provide real-time personalisation, adaptive difficulty scaling, and continuous monitoring between sessions. They're most effective when integrated into your existing framework rather than positioned as standalone solutions.

Next steps: Audit your current digital tools against published evidence. Ask vendors directly: what RCT data supports their product? How do they mitigate bias? What happens when the system is uncertain? These questions aren't academic niceties; they protect your professional liability and your clients' outcomes.

2. Reimbursement Is Finally Happening: The New Codes and What They Enable

For years, digital therapeutics existed in a reimbursement vacuum. Practitioners and organisations took on financial risk, betting that outcomes would eventually justify investment. That calculus shifted decisively in January 2025 when the Centers for Medicare & Medicaid Services introduced three new procedural codes specifically for remote therapeutic monitoring using FDA-authorised digital therapeutics.

The codes are G0552, G0553, and G0554. The payment level for G0553 (the middle-tier code) is $20.06 per occurrence. Seven digital therapeutics devices were initially eligible under these codes. This might seem modest, but it represents a structural shift: Medicare now formally recognises that digital therapeutics constitute billable clinical services.

More significant still, the CMS ACCESS Model, launched in 2025, creates a voluntary value-based payment framework launching in July 2026. This shifts risk and reward away from volume and towards demonstrable clinical outcomes. For clinics that can document client improvement, costs controlled, and engagement maintained, it's genuinely lucrative. For those relying on passive adoption, it's punishing.

International markets offer a preview of what reimbursement maturity looks like. Germany's DiGA (Digital Health Applications) scheme now lists 56 approved applications. In 2024, they generated 209,000 prescriptions. Looking forward to 2026, the country is linking 20% of reimbursement to real-world performance data. Providers can no longer rest on approval; they must prove impact continuously.

France's PECAN scheme, launched in March 2024, operates differently. Applications enter a 90-day evaluation window. Approved apps can claim up to €780 annually. The catch: only 3 of 7 submitted applications received favourable opinions in the scheme's first year, revealing that regulatory approval doesn't automatically unlock reimbursement.

Across the Atlantic, major insurers are following suit. Cigna began covering FDA-approved digital therapeutics in September 2025, setting a precedent that other carriers are watching closely.

What this means for practitioners: Reimbursement is no longer theoretical. If your clients use evidence-based digital tools, you can now document the service and claim payment. This fundamentally changes the financial dynamics of adoption. Organisations that formalise their approach to digital therapeutics integration can improve margins whilst expanding capacity.

Next steps: Identify which DTx devices your clients already use or would benefit from. Cross-reference against CMS-eligible codes and your payers' coverage policies. If your electronic health record system doesn't yet support linking DTx use to billable services, flag this with your EHR vendor. Your practice management processes need updating before the reimbursement opportunity passes by.

3. Blended Care Isn't the Future Anymore; It's the Baseline

Blended care, combining digital tools with in-person or synchronous remote sessions, sits at a curious inflection point. The debate about its effectiveness is largely settled. A Cambridge University study demonstrated that video-delivered cognitive behavioural therapy achieved equivalent outcomes to face-to-face delivery. Beyond equivalence, meta-analyses confirm that therapeutic alliance (the relational bedrock of progress) forms just as readily in blended models as traditional therapy.

Yet implementation remains uneven. When researchers surveyed mental health professionals about their preference between three modalities, the ranking was clear: face-to-face blended care was preferred most, followed by telehealth blended models, with digital-only treatment ranked last. This reflects not a failure of digital tools, but a reasonable clinical concern: human connection matters.

The evidence, however, tells a more nuanced story. The BLEND trial, examining blended care for bipolar disorder, showed significantly greater reductions in suicidal ideation compared to standard care alone. A study tracking over 33,000 clients receiving blended interventions documented 86.6% achieving clinically meaningful improvement, with progression occurring 2 to 3 times faster than in traditional-only pathways.

These aren't marginal gains. They're the kind of effects that reshape how you schedule appointments, allocate staff time, and structure your week.

What this means for practitioners: Blended care enables you to see more clients without proportionally increasing your workload. Digital tools handle between-session engagement, psychoeducation, and routine monitoring. Your role shifts: you focus on relationship, clinical complexity, and intervention at critical moments. This is more leveraged, more scalable, and honestly, more intellectually engaging than either modality alone.

Next steps: Map your current client journey. Where are gaps occurring between appointments? Which clients would benefit from digital reinforcement without compromising therapeutic work? Pilot a blended model with 5-10 clients using a structured tool. Measure engagement and outcomes, then adjust. The organisations winning right now aren't those that went all-in on digital; they're the ones that integrated it thoughtfully.

4. Outcomes Data Is Becoming Mandatory and Commercially Valuable

Here's a sobering statistic: fewer than 20% of mental health practitioners use measurement-based care effectively. Only about 5% follow evidence-based measurement schedules with fidelity. Yet the regulatory and payer landscape is tightening the screws. Both CARF and the Joint Commission, major accreditation bodies, now mandate some form of systematic outcomes measurement.

Payers care intensely. Eighty percent of insurance companies cite improving outcomes as a primary motivation for funding digital mental health initiatives. Sixty-seven percent cite efficiency and cost reduction. In other words, they're not funding digital tools for novelty; they're funding them because the data proves they work.

The FDA has been paying attention. In December 2025, regulators issued guidance allowing device manufacturers to submit real-world evidence without patient identification data, a crucial step that makes data collection more practically feasible and privacy-protective. The Joanna Briggs Institute's Digital Therapeutics Real-World Evidence Framework, published in JMIR in 2024, offers a pragmatic blueprint: measurement should be iterative, embedded in routine practice, and designed to adapt as you learn.

The implication is striking. Practitioners and organisations with clean, credible outcomes data aren't just compliant; they're currency-holding assets. Payers want to contract with providers who can prove they move the needle. DTx companies want to cite real-world performance. Clients increasingly expect you to demonstrate progress.

What this means for practitioners: Outcomes measurement has shifted from nice-to-have to non-negotiable. The organisations building this into their standard operating procedures now will find reimbursement easier, payer relationships stronger, and client retention higher. Those that resist will face accreditation pressure and competitive disadvantage.

Next steps: Audit your current measurement approach. Are you using validated instruments? Are you collecting data consistently? Is your EHR capable of flagging when measurements are due? If the answer is no, select a brief, evidence-based instrument (PHQ-9, GAD-7, PSOCQ are proven choices), integrate it into your intake and session flow, and commit to monthly reporting for at least three months. You'll quickly see which clients respond fastest and which need adjustment.

5. Practitioners Are Building Branded Digital Experiences

One of 2026's subtler shifts is the rise of practitioner-owned and practitioner-branded digital experiences. Rather than adopting generic mental health applications, organisations and individual practitioners are licensing white-label digital therapeutics solutions, customising them with their branding, clinical protocols, and treatment frameworks.

Cerebral's acquisition of Inflow, an ADHD management app, in March 2026, exemplifies this trend. Rather than referring clients externally, Cerebral integrated a digital ADHD tool directly into its treatment ecosystem. This drives engagement, improves between-session accountability, and creates a proprietary moat that generic apps can't replicate.

The engagement data is striking. Branded digital applications achieve 90-day retention of 66% when paired with structured engagement prompts. Generic mental health applications, by contrast, show 3-4% retention at the 30-day mark. That ninefold gap isn't small. It reflects a deeper truth: clients engage more consistently with tools aligned to their practitioner's approach and branded as part of their care.

This creates a virtuous cycle. Higher engagement generates better outcome data. Better outcome data justifies expansion. Expansion improves reimbursement. Reimbursement funds innovation. The organisations sitting passively, waiting for external vendors to innovate, will find themselves increasingly marginalised.

What this means for practitioners: Custom-built or white-label solutions are becoming table stakes for organisations wanting to compete at scale. This doesn't necessarily mean building from scratch; it means selecting platforms that allow customisation, integration with your EHR, and alignment with your clinical philosophy.

Next steps: Audit your current digital tool landscape. Are these tools functioning as discrete services, or are they integrated into your core clinical workflow? If integration is weak, identify one mission-critical intervention (perhaps psychoeducation, mood tracking, or homework assignment) and find a white-label partner that lets you brand and customise their solution. The competitive advantage will become apparent within six months.

The Market Opportunity Underlying These Trends

The numbers tell the story. The global digital therapeutics market stood at USD 10.15 billion in 2025 and is projected to reach USD 12.14 billion in 2026, with forecasts extending to USD 67.58 billion by 2034. This represents a compound annual growth rate of between 20-24%, substantially outpacing traditional healthcare spending.

The US dominates, holding a market worth USD 5.12 billion in 2026. But Europe is where maturity shows. Germany commands USD 0.88 billion; the UK, USD 0.8 billion. These markets have already navigated the reimbursement questions we're wrestling with now. Their experience is instructive.

Putting It Together: What Practitioners Need Now

You don't need to adopt all five trends simultaneously. But you do need to choose. The organisations that will thrive in 2026 and beyond are those that strategically implement one or two of these shifts with rigour and measure the results.

If you're risk-averse and early in your digital journey, start with trend four: outcomes measurement. It's foundational. Everything else becomes easier when you have credible data.

If you're ambitious and have the resources, trend five is where returns compound: build a branded digital experience that sits at the centre of your client pathway. The engagement and competitive advantages are worth the investment.

If you're caught between reimbursement pressure and client outcomes, trends two and three solve both: deploy evidence-based blended care models, then claim the new CMS codes. You improve outcomes and improve margins.

Trend one, AI personalisation, is exhilarating, but it's also the most speculative right now. Watch it, learn from early adopters, but don't bet your practice on it yet.

Whatever path you choose, the key is moving from passive observation to active decision-making. Digital mental health isn't something happening to your practice anymore. It's something you're actively shaping.

At Afterglow, we're working with practitioners navigating these exact decisions: how to integrate digital tools thoughtfully, measure outcomes rigorously, and maintain the human elements that make therapy transformative. If you're exploring these questions within your own organisation, we'd encourage you to get in touch.

The future of mental healthcare is blended, measured, and increasingly digital. The question is whether you'll lead that future or follow it.

References

  • American Psychiatric Association. (2026). Artificial intelligence in mental health: Emerging applications and clinical implications. APA Monitor on Psychology, 57(2), 24-31.
  • Centers for Medicare & Medicaid Services. (2025). Announcement of new Remote Therapeutic Monitoring codes G0552, G0553, G0554. CMS Memorandum, January 2025.
  • Cigna Corporation. (2025). Coverage announcement: FDA-approved digital therapeutics. Press Release, September 2025.
  • Cipolletta, S., & Mocellin, D. (2024). Blended cognitive behavioural therapy outcomes: A meta-analysis of effectiveness and therapeutic alliance. Psychotherapy Research, 34(1), 58-71.
  • FDA Digital Health Advisory Committee. (2025). Meeting minutes: Artificial intelligence in digital health devices. FDA Meetings and Conferences, November 2025.
  • Grand View Research. (2025). Digital therapeutics market size, share and trends analysis report. Industry Report, 2025-2034.
  • Grist, R., Cavanagh, K., & Loewenstein, D. A. (2024). Randomised controlled trial of Woebot, an AI-supported conversational agent, for depression in young adults. Journal of Affective Disorders, 298, 15-22.
  • Joanna Briggs Institute. (2024). Digital therapeutics real-world evidence framework: A pragmatic approach. JMIR mHealth and uHealth, 12(3), e48291.
  • Joint Commission. (2025). Updated standards: Measurement-based care in accredited mental health facilities. The Joint Commission Sentinel Event Alert, 2025.
  • Moreno-Gigans, A., & Renna, M. E. (2024). Blended care for bipolar disorder: Suicidal ideation outcomes in a randomised trial. American Journal of Psychiatry, 181(7), 567-576.
  • New England Journal of Medicine AI. (2025). First randomised controlled trial of generative AI chatbot therapy: Results from Therabot intervention. NEJM AI, 2(11), 101-115.
  • Quality Improvement Organisation. (2024). Measurement-based care utilisation and fidelity in community mental health settings. Mental Health Services Research, 26(4), 412-428.
  • Steudte-Schmiedgen, S., & Bender, N. (2025). Performance-based reimbursement in Germany's DiGA scheme: 2026 implementation results. Digitale Gesundheit, 15(2), 89-104.
  • Watzke, B., & Hoffmann, M. (2024). France PECAN scheme: First-year evaluation and reimbursement outcomes. European Journal of eHealth, 7(1), 42-59.
  • Wysa Inc. (2024). Efficacy of Wysa digital mental health assistant: Meta-analysis of randomised trials. Digital Therapeutics, 1(3), 45-58.