Artificial intelligence has barely reached the direct care workforce so far, but its use is growing, and it is more likely to take on paperwork than to replace hands-on care, according to a new Bipartisan Policy Center (BPC) issue brief.
BPC published "AI's Impact on the Direct Care Workforce" on Oct. 14, 2026. It counts direct support professionals as part of a direct care workforce of 5.8 million.
What the brief says
- Use is low. In the Census Bureau's Business Trends and Outlook Survey, about one-fifth of nursing and residential care facilities said they use AI for anything.
- Where it shows up: home health agencies use AI to assign shifts, help with hiring, process claims and coordinate communication. Some workers use AI-enabled sensors, fall detectors and predictive tools.
- Risks: weak privacy protection, especially for tools outside HIPAA; wrong or incomplete outputs that can cause trouble with Medicaid, Medicare or labor rules; and bias, because older adults, people with disabilities and rural residents are underrepresented in training data.
- Trust gap: nearly nine in 10 direct care workers lack trust that their employers will support them in adopting AI.
BPC says AI won't fix recruitment and retention without better pay, job quality and advancement. It points policymakers to five areas: oversight, public messaging, AI literacy, economic mobility, and supports such as paid leave and child care.
Why this matters for providers and DSPs
For provider agencies weighing AI scheduling or documentation tools, BPC's risk list works as a buyer's checklist: where the data goes, how errors are caught before they reach a billing record or a support plan, and who answers for them. The trust finding suggests bringing DSPs and frontline supervisors into those decisions early, rather than handing them a finished system.