What Will Shape Serious Mental Illness Care in 2026

In recent years, my work as a psychiatrist has expanded to include helping patients sort through an increasingly digital-first and information-heavy world. Misinformation, particularly in a healthcare context, used to suggest a knowledge gap that could be aided through access to more accurate or comprehensive information. What we’re facing today is very different. More frequently, the concern isn’t a lack of information. It’s that the sources or platforms that people are turning to can be incomplete or even incorrect.  As we start to have access to more tools and more data, it’s imperative that we’re thoughtful about how we adopt those into practice.

As a treating clinician, I have seen how essential individual understanding and awareness are to treatment planning, destigmatization and goal-setting. I’m equally aware that the right tools, medications and support systems can be difference-makers. This is just as applicable for clinicians who deserve to grow in their practice approaches alongside advances in care. Looking ahead to 2026, I see several trends that will shape how we care for people living with schizophrenia, bipolar disorder and other serious mental illnesses (SMI).

Prediction 1: Misinformation will continue to complicate mental healthcare

In my practice, I’m seeing more and more patients whose views on treatment have been heavily influenced by sources that are not grounded in psychiatric training, such as social media algorithms, celebrity news, AI-generated chats or unvetted self-help resources.

It is difficult to talk about informed treatment recommendations when certain foundations of shared understanding haven’t been established. Open conversations about diagnosis and treatment are supported by a related awareness of what the illness actually is, what the medications do and how risk and benefit are weighed. Treatment decisions are highly individualized and should consider all appropriate medications including oral and long-acting injectable (LAI) formulations of medications that have been shown through clinical research to be evidence-based treatment options in the real world. 

Introducing LAI medications is often a moment to address misinformation. For example, long-acting antipsychotics like ARISTADA® (aripiprazole lauroxil) are supported by clinical data affirming their safety and efficacy. However, if someone living with an SMI has been led to believe that “injections are dangerous” or “all antipsychotics are sedatives that will change your personality,” providers will first need to understand the origins of those presumptions before introducing additional justification for the treatment plan presented. 

That is a more nuanced conversation, evolving from, “Let’s talk about how this medication may meet your needs” to, “Let’s step back and explore the science of why these medications were developed and the reasons they can be useful in treatment.” It can also require a level of patient-provider trust to delve into the many ways cultures, communities, geographies and generations shape perspectives about healthcare. 

I’m convinced that, in 2026 and beyond, clinicians will increasingly need to be proactive about level-setting with patients, specifically what they do and don’t know, what they believe and where those beliefs come from, before even getting into treatment planning. It’s an extra step, but an important one for shared decision-making.

Prediction 2: Tech and AI can support day-to-day monitoring, but they won’t replace your doctor

AI and digital devices have a valid and growing role in SMI care, supporting helpful objectives like real-time data collection. If we can use an app, a wearable or ecological momentary assessment (EMA) methods to gather real-world information, such as mood ratings throughout the day, sleep patterns, activity levels or heart rate variability, that gives clinicians a more comprehensive and longitudinal evaluation than the scope of what can be covered during an office or clinic visit.

AI tools can help by prompting patients with, for example, reminders to take medications on a specific schedule, nudges to complete screening questions or added check-ins during high-risk periods. That kind of consistent, low-burden monitoring increases touch points between providers and patients in ways that hopefully feel minimally intrusive and are respectful of everyone’s time and autonomy.

However, health tech of any kind, no matter how sophisticated and advanced, isn’t a substitute for the insight and discernment of an experienced human being. When I’m partnering with a patient to manage an SMI, I’m thinking about functionality, relationships, family dynamics, safety, cultural context and how all of those elements change over time, and even from day to day. That’s nuanced, deeply relational work. I don’t think AI is anywhere close to that level of integration and I’m not sure it ever will be.

In 2026, I expect the most responsible care delivery systems will use AI to support both clinicians and patients, generating better and richer data while preserving the essentially human element in diagnosis, treatment planning and therapeutic relationships.

Prediction 3: Through more education and awareness, LAIs can support access in underserved communities

The United States has a measurable access problem in mental healthcare. Across healthcare deserts, which can span rural to urban settings, there aren’t enough clinicians or points of care. Even when providers are available, patients still encounter myriad barriers, including insurance issues, out-of-pocket costs, transportation problems and simply getting to frequent appointments or refilling prescriptions.

Where access to consistent, timely care may be difficult, LAIs can offer a strategic treatment approach. Because LAI administration schedules vary from as infrequently as every two weeks to up to two months or more, logistical demands related to prescription refills or frequent office visits may be lessened. Additionally, patients may find the consistent dosing, with efficacy designed to last over time, lends itself well to symptom management which, in turn, helps with ongoing feelings of stability. Often, as well, LAIs can be initiated in a single day after oral tolerability is determined, making the switch itself from an oral medication relatively convenient.

However, in many underserved areas, providers don’t feel comfortable or have experience prescribing LAIs. I live and practice in Georgia. Most of my colleagues working in the rural areas of the state have had little to no firsthand experience with LAIs. This is not a commentary on their commitment or care as clinicians; it’s a matter of exposure and access. One of the most helpful things we could do in 2026 is provide practical, logistics-focused education for these providers. Not just a theoretical overview, but real guidance on when to consider an LAI, how to move safely from an oral medication to an injectable, and how to handle follow-up and side effect monitoring. Even administration of an LAI in an outpatient psychiatrist’s office can pose a barrier for using this class of medications. Simple, pragmatic training and real-time support can open doors to extremely effective treatment options like the use of LAIs.

Once more clinicians feel confident in their fluency and have direct experience with LAIs, I think they would quickly see how these medications can support treatment goals related to medication adherence, safety and long-term clinical outcomes for their patients.

Prediction 4: Behavioral data will transform how treatment responses are assessed

Two areas where I think technology will genuinely revolutionize care in 2026 are social behavior and sleep.

Social behavior is closely tied to certain components of SMIs. As an example, people living with schizophrenia who are experiencing negative symptoms may default to self-isolation. They spend a lot of time by themselves, at home, not really engaging in conversations or relationships. When they do push themselves to be social, it requires a disproportionate level of exertion, and becomes draining to such a significant degree that many hours spent alone feel like the safer choice. Having objective data about their day, such as how many minutes were spent up and moving around, rather than relying solely on self-reports, would be extremely helpful for treatment planning. We’re very close to being able to use these tools in day-to-day care in a way that respects privacy while delivering meaningful, actionable information.

The same is true for sleep. Sleep patterns are deeply connected to how well psychiatric medications work. When patients sleep better, in my observation it is a good predictor of a positive treatment response for patients with diagnoses such as depression, anxiety and even overt sleep disorders. Having concrete measurements of sleep quality and duration will help providers see how treatments are actually working. It can inform treatment planning decisions as well as conversations about other elements of therapeutic care and psychosocial wellbeing. 

Taken collectively, my 2026 predictions center around the thoughtful, strategic use of many assets we already have access to, and others that are rapidly evolving, to fully leverage accurate information, effective medications, behavioral data and human insight. The ways we think about and treat SMI have evolved considerably in a relatively short amount of time. Looking ahead, I am optimistic we have the tools and data to make measurable progress towards our ultimate goal, improving outcomes for people living with SMI.