Medical staff reviewing data on computers about safe AI use

An Iowa City Medical Practice's Guide To Setting Safe AI Rules

By Dan Roberts on 11/19/26

Your employees may already be using AI at work.

Someone may use it to rewrite an email. Another employee might ask it to summarize information, create patient education materials, brainstorm a social media post, or help with an administrative task.

AI may also be built into software your team already uses. Browser extensions, meeting recorders, transcription tools, and other applications can include AI features employees don't necessarily think of as "using AI."

That leaves many Iowa City medical practices with an important question:

Should we be controlling how employees use AI?

The short answer is yes, medical practices should establish clear rules for employee AI use. That doesn't necessarily mean banning AI. It means knowing which tools employees use, deciding which tools are appropriate, defining what information employees can and cannot share, and making sure people remain responsible for reviewing AI-generated work.

For a medical practice, there's another important consideration:

Patient information should never casually find its way into an AI tool simply because an employee is trying to make a task easier.

Why Medical Practices Need AI Rules

AI can be useful.

The problem is that employees may start using it before leadership has decided how it should be used.

RTS's AI Readiness Guide starts by asking businesses whether they even know which AI tools employees are using. Employees may already be using AI websites, apps, browser extensions, meeting recorders, and built-in AI features without anyone formally approving them.

That's particularly important in a healthcare environment because employees regularly work with sensitive information.

The HIPAA Security Rule establishes safeguards for electronic protected health information (ePHI) and HHS guidance makes clear that covered entities and business associates have responsibilities for protecting this information when using technologies that access it.

AI doesn't make those responsibilities disappear.

So instead of asking:

"Can employees use AI?"

A better question is:

"Under what circumstances can employees use AI safely and appropriately in our medical practice?"

Find Out How Employees Are Already Using AI

Before writing an AI policy, find out what's happening today.

Ask employees which AI tools they use for work and what they use them for.

You may discover employees are using AI to:

  • Draft or rewrite emails
  • Summarize information
  • Create documents
  • Take meeting notes
  • Transcribe conversations
  • Brainstorm ideas
  • Research topics
  • Create patient-facing materials
  • Work with spreadsheets
  • Assist with administrative tasks

You may also discover tools the leadership team didn't realize were being used.

That isn't necessarily evidence employees are doing something wrong. AI adoption has happened quickly, and employees may see these tools as another way to work more efficiently.

The problem arises when every employee independently decides what's appropriate.

Decide What Information Employees Cannot Put Into AI

This is one of the most important parts of an AI policy for a medical practice.

Employees need clear instructions about what information is off-limits.

RTS's AI Readiness Guide tells businesses to define what can and cannot be shared with AI and specifically calls attention to customer information, financial data, passwords, internal documents, and other sensitive information.

For a medical practice, patient information requires particular care.

HHS says regulated healthcare entities must protect ePHI and that disclosures of PHI to technology vendors must comply with applicable HIPAA requirements. Whether a particular AI vendor can appropriately handle PHI depends on the specific tool, how it's being used, the relationship with the vendor, and other circumstances.

That's why an employee shouldn't assume it's okay to copy patient information into an AI tool just because the tool is easy to access.

Identify Which AI Tools Are Approved

"AI" isn't one product.

There are public AI websites, paid business platforms, browser extensions, AI assistants built into software, transcription tools, meeting recorders, and countless specialized applications.

They don't all handle information the same way.

RTS recommends businesses decide which AI tools, apps, extensions, meeting recorders, and integrations are approved.

For a medical practice, approving a tool should involve more than asking whether employees like it.

The organization needs to understand what information the tool accesses, where that information goes, how it's handled, and whether its use is appropriate for the practice.

If the AI tool will create, receive, maintain, or transmit PHI for a healthcare provider, it must follow HIPAA rules. HHS specifically lists third-party AI chatbots that handle patient PHI as business associates. This means the medical practice needs a signed Business Associate Agreement (BAA) before using chatbots.

Don't Forget About AI Meeting Recorders And Browser Extensions

These tools are easy to overlook.

An employee may think:

"It just takes notes during meetings."

Or:

"It's only a browser extension."

The important question is what information the tool can access.

RTS's AI Readiness Guide specifically warns businesses to consider AI meeting recorders, browser extensions, and automation because these tools can potentially have extensive access to email, meetings, files, and company systems.

That matters in a medical practice.

A meeting recorder that captures a conversation containing patient information is very different from an AI tool used to brainstorm ideas for the company picnic.

The technology may be convenient.

The information being exposed to it still matters.

Require Employees To Check AI-Generated Work

AI can produce impressive work.

It can also be wrong.

Sometimes it can be wrong while sounding completely confident.

That's why RTS includes "Verify before you trust" among its 10 AI Safety Rules and advises businesses never to let AI make the final decision.

For medical practices, human review is particularly important.

Employees shouldn't assume something is accurate simply because AI generated a polished answer.

AI can assist with work.

The person using it still needs to be responsible for the result.

Teach Employees When To Stop And Ask

No AI policy will anticipate every situation.

A front desk employee may encounter a tool leadership has never seen.

A manager may discover a new AI feature inside software the practice already uses.

Someone may have a document they aren't sure is appropriate to upload.

Employees need to know what to do when the answer isn't obvious.

RTS's tenth AI Safety Rule is deliberately simple:

"When in doubt, stop and ask."

That can be one of the most valuable rules in the policy.

Employees shouldn't have to make cybersecurity, privacy, or compliance decisions on their own.

Put The Rules In Writing

Simply telling employees to "Be careful with AI," is vague, leaving room for a wide interpretation of "being careful."

One employee may think using AI to rewrite an email is harmless. Another may assume uploading a document is fine. Someone else may install a meeting recorder because a colleague recommended it.

A written AI Acceptable Use Policy creates a common standard.

At minimum, a policy can address:

  • Which AI tools are approved
  • Which tools or uses are prohibited
  • What information must never be entered
  • When human review is required
  • Whether employees can install AI extensions or integrations
  • How meeting and transcription tools should be handled
  • Who employees should ask when they're unsure

RTS's AI Readiness Guide makes this one of its core readiness questions: Does your business have a written AI Acceptable Use Policy? The guide recommends clear, consistent rules about what's allowed, what's prohibited, and what employees should do when they're uncertain.

Should We Just Ban AI?

For some organizations, prohibiting certain tools or uses may be appropriate.

But "ban AI" and "let everyone use whatever they want" aren't the only choices.

AI features are increasingly appearing inside ordinary business applications. A blanket prohibition may also fail to address the larger question of how the organization intends to evaluate and manage these tools.

A more useful approach is to make intentional decisions.

Which tools do we approve?

What are employees allowed to do with them?

What information stays out?

Where is human review required?

Who decides when a new tool is appropriate?

Those questions create governance rather than guesswork.

Who Should Be Responsible For AI Use In A Medical Practice?

AI shouldn't become solely an IT issue.

Technology professionals can help evaluate tools, security, access, integrations, and data-handling concerns.

Practice leadership needs to decide how AI fits into operations and what level of risk is acceptable.

Employees need training and clear expectations.

And when AI use raises legal, regulatory, clinical, contractual, or compliance questions, the appropriate professionals should be involved.

The goal isn't to make one person the "AI police."

It's to establish accountability.

Frequently Asked Questions

Should medical practices allow employees to use AI?

Medical practices can establish approved uses for AI rather than leaving every employee to make the decision independently. Policies should identify approved tools, prohibited uses, information that shouldn't be shared, human-review requirements, and what employees should do when they're unsure.

Can employees put patient information into ChatGPT or another AI tool?

Employees shouldn't assume patient information can be entered into an AI tool. When PHI is involved, HIPAA requirements may apply to how the information is used or disclosed and to the relationship with the technology vendor. The specific tool and use case need to be evaluated.

Does an AI tool need to be HIPAA compliant?

"HIPAA compliant" shouldn't be treated as a simple product label that answers every question. When a vendor creates, receives, maintains, or transmits PHI on behalf of a regulated healthcare entity, HIPAA requirements can apply, including business-associate requirements in appropriate circumstances. The organization still needs to evaluate how the technology is actually being used.

Should employees use AI to make patient care decisions?

AI use involving clinical decisions requires careful evaluation beyond a general workplace AI policy. RTS's broader guidance is that AI shouldn't make the final decision and that important AI-generated information requires human review.

What should be in a medical practice AI policy?

A policy can identify approved tools, prohibited uses, rules for sensitive information, human-review requirements, employee responsibilities, and a process for asking questions or approving new tools.

What if employees are already using AI without approval?

Start by finding out which tools they're using and why. Then determine which uses are appropriate, identify potential risks, establish approved tools, and give employees clear rules going forward.

How often should we review our AI policy?

AI tools and their capabilities change quickly. Review your policy when your technology, tools, business processes, or relevant requirements change rather than assuming a policy written today will remain appropriate indefinitely.

Start By Finding Out How Ready Your Practice Is

You don't need to become an AI expert before setting reasonable rules for your Iowa City medical practice.

Start with a few questions.

  1. Do you know which AI tools your employees use?
  2. Have you decided what information can and cannot be shared?
  3. Do employees know which tools are approved?
  4. Does someone verify AI-generated work?
  5. Have you addressed meeting recorders and browser extensions?
  6. Have employees received training?
  7. Do you have a written AI Acceptable Use Policy?

Those are the same kinds of questions included in RTS Technology Solutions' AI Readiness Guide. The guide includes 10 AI Safety Rules, a readiness check, a scorecard, and practical next steps for businesses establishing AI safeguards.

The goal isn't to scare employees away from AI.

It's to give them enough direction to use new technology without making decisions about sensitive information and business risks on their own.

Get the Free AI Readiness Guide from RTS Technology Solutions by filling out the form at the top.

If your readiness assessment uncovers gaps, RTS can also help your practice think through approved tools, data protection, employee expectations, and an AI Acceptable Use Policy.

Talk With RTS About AI Strategy