Chances are someone at your agency has already used AI at work this month. Maybe your scheduler asked a chatbot to tidy up a Client’s visit notes, or your office manager had it draft a Caregiver recruiting ad in half the usual time. Those shortcuts save real hours, but most agencies haven’t decided which tools are approved or what Client and Caregiver information is allowed to go into them.
To help you set those rules before a problem forces the issue, we asked Angelo Spinola, Home Health, Home Care & Hospice Chair at Polsinelli, where AI fits in an agency’s daily work and where it creates legal and privacy risk. With two decades of employment law experience focused on home-based care providers, Angelo brings a practical view that’s easy to share with your office team this week. 
1) What are the biggest risks agencies take on when they adopt AI tools without a clear policy in place?
Using AI tools without a clear policy can lead to privacy issues, including improper use of client or caregiver information, and violations of HIPAA or state privacy rules because people handle data differently. It can also cause day-to-day mistakes if staff trust AI answers that haven’t been checked, use tools the agency hasn’t approved, or aren’t sure who is responsible—hurting the agency’s reputation and increasing legal and contractual risk.
2) Where do you see AI adding the most value for home care agencies right now, and where should agencies still be cautious?
AI can help with drafting care plans and administrative paperwork, client scheduling, remote monitoring, claim and reimbursement submissions, route optimization, and supporting staff training. Be careful using AI for anything that guides clinical care, rates a client’s risk, or affects care plans, staffing decisions, or billing, which all require human oversight from a person with subject matter knowledge. Do not enter private client or caregiver information into tools that haven’t been approved and secured.
3) What should an agency’s AI use policy cover, at a minimum, to protect both the business and its clients?
The policy should clearly list which AI tools staff may use, what they may not use them for, what kinds of client and caregiver information can be used, and basic rules for handling that information (what can be entered, when to remove identifying details, and how long information is kept). It should also require a person to review AI output, set expectations for double-checking accuracy and fairness, spell out how vendors are approved and kept secure, explain how to report problems or data leaks, limit access to those who need it, keep records for review, train staff, and state what happens if the policy is not followed.
4) Are there compliance or privacy considerations that are unique to home care when it comes to AI, especially with client and caregiver data?
Client information can include health details, notes about behavior or the home setting, and location data, so it should be used only after vetting for HIPAA compliance, kept secure when shared or stored, and covered with the right terms in vendor agreements. Caregiver information (schedules, location tracking, performance notes) should be collected and used only for real business needs, with clear notice to staff and permission where required, and with checks to make sure automated tools don’t unfairly affect hours, assignments, or pay.
5) What’s one mistake you commonly see agencies make when rolling out a new AI tool?
Agencies often roll out a new AI tool without testing it first, training staff, or setting clear rules, which can lead to inconsistent use, accidental sharing of private information, and less confidence in the results. A better approach is to start with one clear purpose, confirm privacy requirements up front, have a person check the AI’s work, and track whether it’s accurate and actually helping—so you avoid rework and compliance problems.
Angelo’s advice keeps pointing back to the same starting place: pick one task for AI to handle, and write down the rules for it before anyone on your team starts using the tool. That means deciding which Client and Caregiver details stay out of it and who reviews the output before it affects a care plan or a paycheck.
For a deeper look at how AI is affecting the industry, Polsinelli’s latest Home Care Industry Quarterly Update includes a section dedicated to the topic. Thank you to Angelo and the Polsinelli team for sharing their expertise with Rosemark readers.