Nashville, TN — The Gaylord Opryland Resort and Convention Center welcomed executives, clinicians, and technology leaders from across the behavioral health industry for HMP Global's inaugural Behavioral Health AI Summit on April 7–8. The event gathered an audience of C-suite and senior behavioral health operations leaders who are all navigating one of the most consequential questions facing their organizations right now: How will AI reshape behavioral healthcare—and how can their teams take the necessary steps to prepare to incorporate AI solutions to be in their workflows?
I spent time on the exhibit floor and in the sessions engaging with behavioral health leaders at varying stages of their AI journeys. Here are three themes that defined the conversation:
1. Most organizations are still finding their footing—and that’s okay.
The Behavioral Health AI Summit drew a room that was refreshingly candid about where it actually stands with AI. Many attendees were in the early stages of exploration: curious, cautious, and still developing the internal frameworks they'll need before committing to broader implementation. Others came as enthusiastic advocates for AI adoption, looking for practical roadmaps for responsible integration across clinical and operational workflows.
The first day’s keynote speaker, author Isamu Pant, asked the audience to define their stage of adoption:
- AI FOMO: Those who are experiencing “FOMO” and are eager to explore AI applications
- Shadow AI: Those who are silently using chatbots, transcription tools, and other AI resources unbeknownst to their organization
- Demo-rich, outcome poor: Those who have seen demos for AI solutions but are unsure of how to evaluate them
- Force multiplier: Those who have successfully implemented AI systems to achieve meaningful improvements to clinical workflows
The categories resonated.
Pant then reframed these stances through a governance lens: What AI policies and guardrails does your organization have in place? Were organizations aware of who has already implemented AI solutions and how? Pant’s message was clear: the first step toward meaningful AI adoption is knowing where the gaps in your own organization's readiness actually are.
I've had versions of these same conversations with our customers at Charta. The behavioral health organizations we work with are often similarly positioned: they’re not resistant to AI, but they’re trying to be appropriately thoughtful about what it means to deploy it in an environment defined by complex regulations, unique patient populations, and documentation requirements that carry real clinical and financial consequences. It’s one of the main reasons we customize every client integration.
2. Smarter documentation shouldn’t mean weaker clinicians.
One of the most salient sessions of the conference—"AI in the Patient Journey: AI and Clinical Tools and Documentation"—also posed an important question to attendees: Are AI applications creating smarter documentation but weaker clinicians?
The answer suggested by Dr. Cara Bishop and Dr. Rachel Jackson was that this doesn’t have to be the outcome: AI documentation tools should empower clinicians, and a key element of that empowerment is more dynamic chart reviews that provide timely and detailed provider feedback.
This is precisely where AI chart review becomes a clinician's partner rather than an overseer. When AI systematically reviews 100% of charts to flag documentation gaps, payer-specific compliance issues, and coding inaccuracies, it functions as a continuous feedback loop for clinical staff. Clinicians learn where their documentation patterns create downstream problems. They don't just produce better notes: they develop a deeper understanding of what payers require, what state regulations demand, and what constitutes clinical documentation integrity in behavioral health.
3. The industry is ready to move from pilot to platform—but needs a clear path.
On day two, the conversation turned toward scaling: How do you move from a promising AI pilot to an embedded operational platform that delivers real results? What separates organizations that have successfully expanded AI from those stuck in perpetual experimentation?
Day Two presentations, including a case study from Community Medical Services on integrating AI across multiple sites while preserving care quality and engagement, attempted to separate proof-of-concept from proof-in-outcomes. A well-thought-out AI implementation requires planning across different areas: governance, stakeholder alignment, and technology. This cross-functional approach flexes an organization's full operational footprint to ensure a new implementation isn’t needed every time the regulatory or clinical environment changes.
For Charta, the “Pilot to Platform” conversation is one we engage in every day. AI chart review doesn't work as an AI solution bolted onto an existing process. It works because it's embedded deep into the revenue cycle as a continuous, systematic review of every chart across clinical dimensions, flagging exceptions before they become denials, surfacing compliance risks before they become penalties, and generating the kind of performance data that lets RCM and clinical leadership make collaborative and informed decisions about where and how to improve outcomes. That's what it means to move from pilot to platform: the technology becomes infrastructure, not a project.
The Summit raised a lot of questions, but made one thing clear: behavioral health leaders aren't waiting to be convinced that AI matters. They're working through the harder questions of how to deploy it responsibly, how to build the internal capacity to use it well, and how to ensure that the efficiency gains don't come at the expense of the human relationships that define behavioral healthcare. Those are the right questions. And the answers, increasingly, start with better visibility into what's actually happening at the chart level.