Building AI in Neurodiversity Care the Responsible Way

March 2026 | Eve Mamane, Co-Founder at Agave Health

Neurodiversity care is at an inflection point.

In 2025, the CDC estimates that approximately 15.5 million U.S. adults have a current ADHD diagnosis, representing roughly 6 percent of the adult population, with prevalence continuing to rise (CDC, 2025). Nearly 1 in 5 U.S. adults lives with a mental health condition (NIMH, 2024), and provider shortages persist across most states (HRSA, 2024). Demand is accelerating faster than the traditional care system can absorb.

Artificial intelligence is entering healthcare just as quickly. More than 60 percent of healthcare organizations report active generative AI initiatives (McKinsey, 2024), and investment into AI-driven healthcare companies continues to scale (CB Insights, 2024).

In neurodiversity care, however, speed cannot outrun responsibility.

At Agave Health, we believe AI belongs in neurodiversity care. We also believe it must operate inside clinical, regulatory, and ethical guardrails that reflect the realities of healthcare.

Adults with ADHD are navigating executive dysfunction, emotional volatility, workplace pressure, and often years of missed or misunderstood symptoms. The need is not simply for more content or faster tools, but for structured, reliable support that understands the complexity of how neurodivergence shows up in real life. This is clinical territory. It requires rigor and trust.

Our position is straightforward: AI should augment care, not impersonate it.

Within Agave, AI enhances structured reflection, supports goal clarity, and helps organize information so that care becomes more precise and continuous. It does not diagnose. It does not prescribe. It does not independently determine treatment pathways. Human oversight remains central, with licensed professionals retaining authority over care decisions and AI functioning as a supportive layer within defined boundaries.

This approach is especially important in employer-sponsored healthcare, where most Americans receive coverage (KFF, 2024). Trust is fragile in these environments. Individual-level health data is not visible to employers. Reporting is aggregated and de-identified. Protected Health Information is not used for advertising or generalized model training.

Agave operates under HIPAA-aligned safeguards and is compliant with SOC 2 Type II standards. Our AI infrastructure functions under signed Business Associate Agreements, including with OpenAI, with contractual restrictions governing data use. These are architectural decisions, not marketing language.

The capabilities of large language models in healthcare are expanding rapidly, and so are the conversations about their limitations. Research has documented both their promise and the need for careful implementation, particularly in sensitive clinical contexts (Nature Medicine, 2023). In neurodiversity care, where nuance matters and individual trajectories vary significantly, AI must operate within clearly defined scopes and safety frameworks.

As access models evolve, AI can responsibly expand reach, including through structured, supervised digital coaching experiences designed to increase affordability and continuity. Expanding access does not require lowering standards. It requires building systems that are transparent about their role, bounded in their scope, and accountable in their operation.

The future of ADHD and neurodiversity care will be hybrid. Digital systems can increase continuity, personalization, and accessibility. Human professionals provide judgment, interpretation, and therapeutic alliance. The strongest models will integrate both.

Neurodiversity care deserves durability, not hype.

At Agave Health, we are building healthcare for the long term. And in healthcare, trust compounds.

Read more about Agave Health's AI position here.