TL;DR
Turn quiet afternoons into listening time
- Thousands of audiobooks, podcasts and originals
- Listen on your phone, tablet or Echo — also offline
- Cancel anytime
Home-based care providers are exploring agentic AI for scheduling and administrative tasks, but adoption remains cautious because agents can take actions and make mistakes with real operational consequences. TheKey is rolling out an AI scheduling tool and developing a broader strategy, while legal and industry leaders urge providers to assess risks and protect care delivery.
Home-based care providers are testing agentic AI for scheduling and other administrative work, while legal and industry leaders warn that autonomous systems could create patient-safety and compliance risks if they make mistakes. In a report by Home Health Care News, TheKey described plans to expand AI-agent support across employee roles, while Modivcare said it is prioritizing operational improvements and proceeding cautiously.
Agentic AI differs from conventional generative AI because it can interpret goals, plan steps and use tools to carry out actions with some autonomy, according to Stanford University’s definition cited in the report. That ability may let agents handle tasks such as finding a replacement caregiver after a shift cancellation, but it also means an error can affect operations rather than simply produce an incorrect text response.
TheKey is rolling out Caribou’s Conductor scheduling product, which the company expects to help find coverage when a caregiver calls out. Chief People Officer Emily Kraeter told Home Health Care News that TheKey is working toward a model in which employees in different roles have two or three AI agents supporting them. The company says it is examining employee workflows to identify where agents could save time, including for caregiver managers.
Modivcare is taking a more incremental approach, according to CEO of personal care services Marshalina Ramos. She said agents could help streamline call queues, but described the company as focused on lower-hanging opportunities to improve operational efficiency. The report does not establish that the providers’ pilots have produced measured savings or improved care outcomes.
How AI Errors Could Affect Care
Agentic systems can act on workflows, so their mistakes may have consequences beyond inaccurate information. Polsinelli attorney Greg Leighton told Home Health Care News that a hypothetical patient-monitoring agent could miss a serious medical issue or misjudge its urgency. He said the consequences in home health could be severe. This was an example of a possible risk, not a reported incident involving a provider’s deployed system.
Administrative uses may carry different levels of risk from tools that influence clinical monitoring or decisions. Leighton identified operations, appointment coordination, billing and reimbursement management as possible uses, while urging providers to understand the business case and ways to limit downsides before deployment. For agencies, the practical question is whether a tool frees staff to support caregivers without disrupting patient care or creating new compliance problems.
Adoption also has consequences for workers and service reliability. TheKey says AI support could give caregiver managers more time for training, relationship-building and employee support. Those are company expectations; the report provides no outcome data confirming that the scheduling product has achieved them.
Popular right now
Dyson V15 Detect Cordless Vacuum
Samsung 48 4K OLED Smart TV (2026) with WireLogic 8ft Ultra High Speed HDMI Cable
Shark AI Ultra Robot Vacuum with Matrix Clean Navigation, Self-Empty Base, HEPA Filtration, Pet Hair Friendly, Wi‑Fi, Dark Grey
As an affiliate, we earn on qualifying purchases.
From Scheduling Pilot to Strategy
TheKey’s AI effort began with automated scheduling, according to Chief Operating Officer Greg Meyers. He told Home Health Care News that the work prompted the company to consider what agents might do across the organization, including in his own role. TheKey provides home care and memory care, among other services, across 29 states and Canada, and is working to expand its home health services.
Modivcare operates across several care-related services, including nonemergency medical transportation, personal care, remote monitoring and integrative supportive care in 35 states. Ramos said the company sees potential for agents to streamline call queues but is not seeking to be first to market. The report describes providers as split over the pace of adoption, with possible efficiency gains weighed against the demands of a highly regulated sector.
Leighton recommends a staged approach: use generative AI or other non-agentic tools to work through existing processes and compliance issues before moving to workflows where agents act more independently. That is his advice, not a reported industry rule or a formal regulatory requirement.
“When you go to agentic, you’re really putting your operational risk in the hands of the tool.”
— Greg Leighton, privacy and incident response vice chair at Polsinelli Law Firm
Pilot Results and Safeguards
The report does not give deployment dates, pilot performance figures or measured outcomes for TheKey’s scheduling product. It also does not detail how the company will review agent actions, handle errors, protect sensitive information or decide when a human must intervene. Those safeguards are especially relevant as providers consider uses beyond back-office tasks.
It is also unclear how broadly providers will adopt agentic AI or which specific applications will move from testing into routine use. Ramos cited fraud, waste and abuse concerns and federal enforcement activity as reasons providers are cautious, but the report does not describe a new rule or enforcement action tied specifically to these AI tools. No patient harm from the providers’ described projects was reported.
Provider Tests and Oversight
TheKey’s next step is to roll out its scheduling product and assess where AI agents may save employees time, according to the company’s account in Home Health Care News. The report does not provide a public timetable or specify how results will be measured. Modivcare’s comments point to a continued focus on operational-efficiency opportunities rather than an immediate push for broad agent deployment.
Providers considering agents will need to judge each task on its own risks, including the potential impact of an incorrect action and the oversight required to catch one. Leighton’s advice is to establish the business case and manage downsides before proceeding, while Ramos said new technology must not compromise care. Whether early scheduling and administrative uses can meet those conditions remains to be seen.
Key Questions
What is agentic AI?
Agentic AI can plan and take actions to pursue a goal, using tools with some autonomy. Stanford University’s definition cited in the report distinguishes it from generative AI systems that respond to prompts but do not independently complete tasks.
How is TheKey using agentic AI?
TheKey is rolling out Caribou’s Conductor scheduling product, which it expects to help find caregiver coverage when someone calls out. The company is also considering AI-agent support for employees in different roles.
What risks do providers see?
Leighton warned that an agent could make an operational mistake, with greater potential consequences if it affects patient monitoring or care coordination. Providers also face compliance concerns in a heavily regulated industry. The report did not describe a specific harmful incident involving these projects.
Have the AI tools shown improved outcomes?
The report provides no measured results for savings, staffing, care quality or patient outcomes from the described projects. TheKey and Modivcare discussed potential uses and plans, rather than publishing performance data.
Source: homehealthcarenews.com
