AI Safety Training for Healthcare Teams

AI Safety Training for Healthcare Teams

Your Healthcare Team Is Already Using AI. Have You Trained Them to Use It Safely?

The Hidden Risk: Employees Are Using AI Without Training

“This AI tool is amazing!”

Imagine this:

One of your clinic employees discovers a new whiz-bang AI tool that promises to slash their workload.

It’s easy to use, the results look impressive, and best of all, there’s a free version!

The employee decides to try it out and uploads a spreadsheet to help organize staff schedules.

Then someone asks, “Did you just share our employee list with AI?”

Oops.

The employee hadn’t considered that uploading the spreadsheet might disclose personal information to an external service provider.

There was no malicious intent. The employee was simply trying to be more efficient.

This is exactly why AI safety training for healthcare teams matters.

Employees may be comfortable using AI tools without understanding the privacy, security, accuracy, and professional responsibilities that come with them.

AI Literacy Is Not the Same as AI Safety

Many healthcare employees already use AI in their personal lives. They may use ChatGPT to draft messages, research topics, or organize information.

However, using AI at work introduces additional responsibilities.

Healthcare employees need to understand:

  • Privacy and confidentiality: What patient, employee, and business information must not be entered into unapproved AI tools?
  • Approved AI tools: Which applications are authorized by the clinic, and for what purposes?
  • Accuracy and reliability: How can AI generate convincing but incorrect information?
  • Human oversight: Who is responsible for reviewing and verifying AI-generated content before it is used?
  • Transparency: When should patients and colleagues be informed that AI is being used?
  • Accountability: What should employees do if they accidentally share confidential information with an AI application?

AI safety training helps employees recognize these risks before a privacy breach, clinical error, or other incident occurs.

AI Training for Healthcare Is Part of Your Privacy Management Program

In Canadian healthcare practices, protecting personal health information is already a fundamental responsibility.

Introducing AI doesn’t remove those obligations. It creates new situations in which employees must apply them.

For example, an employee might use an AI tool to summarize a patient complaint, draft a referral letter, or analyze clinic operations.

Even if the employee believes the information has been de-identified, it may still contain details that could identify an individual.

Your clinic’s privacy and security policies should address AI use. But policies alone aren’t enough.

Employees need practical training to understand what those policies mean in their day-to-day work.

AI awareness training should be included in your onboarding program, reinforced through ongoing privacy awareness activities, and updated as your clinic introduces new AI tools.

Don’t Let Employees Bring Their Own AI to Work

One of the messages I emphasize in my privacy consulting work is:

Implementing AI in a healthcare practice is a business and custodian decision—not an individual staff decision.

The clinic’s leadership and responsible health information custodian must determine whether an AI application is appropriate for the practice and ensure that applicable privacy, security, and professional requirements are addressed.

Before implementing an AI application, the clinic should assess the vendor, intended uses, privacy and security safeguards, contractual requirements, and any applicable privacy impact assessment obligations.

This applies to free AI tools, paid subscriptions, AI scribes, AI receptionists, and other applications.

It also applies to locums, contractors, and temporary employees.

Employees should not independently introduce AI tools into clinic workflows simply because they are convenient, free, or promise to save time.

AI safety training helps employees understand who can authorize AI tools, which tools they are permitted to use, and when they must seek approval before trying something new.

AI Scribe onboarding package for locums infographic

Start With AI Safety Training for Healthcare Teams

You don’t need every employee to become an AI expert.

You do need everyone to understand the fundamentals of safe and responsible AI use.

A good starting point is practical, role-appropriate training that explains how AI works, what it can and cannot do, common risks, and when to ask questions before using it.

This gives your team a foundation for making better decisions, recognizing potential problems, and using approved AI tools more confidently.

The goal is not to discourage AI adoption. It is to help your team use AI safely, responsibly, and with appropriate human oversight.

AI Safety Essentials for Healthcare Teams

During my October 13 Q&A with Jean, I welcomed Anne Genge of Myla Training to discuss why AI safety training is important for healthcare teams.

Anne shared practical insights about AI awareness, the risks of unapproved AI use, and the importance of preparing employees before introducing AI into healthcare workflows.

I also completed Myla Training’s AI Safety Essentials for Healthcare Teams course myself.

I found the course practical, well-presented, and easy to follow, with relevant examples for healthcare. I particularly appreciated the module on AI agents, which made a complex topic easier to understand.

I recommend this course to healthcare clinic owners, managers, privacy officers, and employees who want a better understanding of how to use AI safely at work.

Take the Next Step: Train Your Team on AI Safety

Don’t assume your employees already know how to use AI safely.

Give them the knowledge and confidence to recognize AI risks, protect confidential information, and make informed decisions.

Explore Myla Training’s AI Safety Essentials for Healthcare Teams course and receive a 10% discount using my referral link.

Affiliate disclosure: This is my affiliate link. You receive a 10% discount when you use the link, and I receive a small commission. There is no additional cost to you.

​You Might Also Be Interested in These Articles…

AI Scribes in Canada
Learn about AI scribe initiatives in Canada, AI governance, privacy impact assessments, and the importance of approving AI tools before introducing them into your healthcare practice.

Thinking About Using AI Scribe in Your Healthcare Practice?
Explore the benefits and privacy considerations of AI scribes, including patient notification, vendor selection, and responsible implementation.

Your Locum Has an AI Scribe. Is It Approved for Use in Your Clinic?

Your Locum Has an AI Scribe. Is It Approved for Use in Your Clinic?

Your Locum Has an AI Scribe. Is It Approved for Use in Your Clinic?

Prevent Shadow AI: Make Sure Locums Follow Your Clinic’s Privacy Policies and Procedures

You’ve arranged for a locum healthcare provider to cover your practice. Great! You’ve organized the schedule, prepared EMR access, and made sure everything is ready.

But have you asked one important question?

“Are you planning to use an AI tool while working in our clinic?”

Your locum may already be using an AI scribe. They may love it! It saves time, reduces documentation workload, and helps them focus on patient care.

But just because the locum is familiar with an AI tool doesn’t mean that it has been approved for use in your clinic.

And that’s where privacy risks can begin.

What Is Shadow AI?

Shadow AI happens when employees, contractors, or healthcare providers use artificial intelligence tools for work without the organization’s knowledge or approval.

Imagine a locum who arrives at your clinic with their favourite AI scribe already installed on their mobile phone.

They begin recording patient encounters and generating clinical notes, assuming it’s acceptable because they’ve used the application elsewhere.

But your clinic hasn’t reviewed the AI vendor, completed the necessary privacy assessment, or confirmed where the patient information is stored and processed.

You haven’t provided patients with the appropriate notice about AI use or obtained their acknowledgement.

Perhaps the audio recordings remain on the locum’s personal device. Perhaps the vendor retains information longer than your clinic permits.

Now you have a potential privacy breach—and a problem that could have been prevented with better preparation.

Custodians: Your Clinic’s Rules Still Apply

Locums are valuable members of your healthcare team. They bring experience, flexibility, and much-needed coverage.

But they also need to understand and follow your clinic’s privacy and security policies and procedures.

As a custodian responsible for health information, you need to establish clear expectations about how patient information is collected, used, disclosed, stored, and protected.

That includes the use of AI tools.

You don’t have to prohibit AI. You do need a process to determine which AI applications are appropriate for your clinic and how they may be used.

If a locum wants to use an AI tool that your clinic hasn’t approved, there should be a process to request a review before the tool is used with patient information.

AI Scribe onboarding package for locums infographic

Prepare an AI Onboarding Package for Locums

One of the easiest ways to prevent shadow AI is to prepare an AI onboarding package that you can share with every locum before they begin working in your practice.

Your package should include:

1. Your AI Governance Framework

Explain your clinic’s rules for adopting and using AI tools. Identify who approves new AI applications, what assessments are required, and the responsibilities of healthcare providers.

2. Your List of Approved AI Tools

Identify approved applications and explain that personal AI accounts and unapproved tools must not be used with patient information.

3. Oath of Confidentiality

Ensure the locum acknowledges your clinic’s confidentiality, privacy, and security requirements. Professional experience doesn’t replace the need to follow your organization’s policies.

4. AI User Training and Acknowledgement

Don’t assume that because someone knows how to use an AI application, they know how to use it safely in your clinic.

Provide practical instructions about patient notices and consent where applicable, audio recording and deletion, transferring AI-generated notes into the EMR, and reviewing AI-generated content.

5. Monitoring and Follow-Up

Your responsibility doesn’t end when the locum signs the acknowledgement.

Confirm that approved tools are being used appropriately. Address questions, reinforce expectations, and document your follow-up.

Remember: AI Doesn’t Replace Professional Judgment

An AI scribe may generate a draft clinical note, but the healthcare provider remains responsible for reviewing the documentation for accuracy, completeness, and appropriateness before it becomes part of the patient record.

AI-generated documentation can contain errors, omissions.

Your AI policies and training should make that responsibility clear.

Don’t Wait Until the Locum Arrives

Imagine how much easier this conversation will be if you have already prepared your AI onboarding package.

Before the locum’s first day, you can explain:

“Here are the AI tools we’ve approved for use in our clinic. Here’s how we use them safely. If you would like to use a different application, please talk to us first.”

That’s a much better conversation than discovering an unapproved AI tool after patient information has already been shared.

Your goal isn’t to make life difficult for locums. It’s to help them provide excellent patient care while protecting the privacy and security of your patients’ information.

Your Next Step: Download the AI Governance Framework

Do you have an AI Governance Framework for your healthcare practice?

If not, now is a good time to get started.

An AI Governance Framework helps your clinic establish clear expectations for approving, implementing, and monitoring AI tools—including those used by locums, contractors, and employees.

Download the AI Governance Framework Template to help you develop practical rules for responsible AI use in your healthcare practice.

Get Your AI Governance Framework Template​

​You Might Also Be Interested in These Articles…

AI Scribes in Canada
Learn about AI scribe initiatives in Canada, AI governance, privacy impact assessments, and the importance of approving AI tools before introducing them into your healthcare practice.

Thinking About Using AI Scribe in Your Healthcare Practice?
Explore the benefits and privacy considerations of AI scribes, including patient notification, vendor selection, and responsible implementation.

Privacy Officer Check-Up 10 Things You Should Review

Privacy Officer Check-Up 10 Things You Should Review

Privacy Officer Check-Up: 10 Things You Should Review

When was the last time you checked whether your privacy management program still reflects how your healthcare practice actually operates?

Employees change. Technology changes. Vendors change. New AI tools appear. Meanwhile, policies, procedures and Privacy Impact Assessments can quickly become outdated.

A Privacy Officer Check-Up helps you recognize what you have accomplished, identify what has changed and prioritize what needs attention next.

You do not need to fix everything today. The goal is to know where you are—and what comes next.

1. Confirm Your Privacy Officer Information Is Current

Check that the correct Privacy Officer information appears in your policies, collection notices, patient information, website, Privacy Impact Assessments and information provided to your privacy regulator, where applicable.

Identify the Privacy Officer including a telephone number or email address patients can use.

Most importantly, make sure everyone on your team knows who the Privacy Officer is and how to contact them.

2. Review Privacy Awareness Training

Privacy training should be included in onboarding, regular annual refresher training and updates following changes to legislation, technology, policies or privacy risks.

Registering someone for training is not the same as confirming they completed it. Your records should demonstrate who completed the training and when.

3. Review Oaths of Confidentiality Agreements

Confirm that custodians and affiliates—including employees, contractors, students and volunteers—have signed current and appropriate Oaths of Confidentiality.

Check that the agreements are signed, current, appropriate to each person’s responsibilities and accessible in the applicable personnel or contractor record.

4. Review User Accounts and Access Permissions

Ask your system administrators or IT provider to generate current user-account reports for your:

  • EMR, pharmacy management or clinical systems
  • Email and business computer systems
  • Microsoft 365 or Google Workspace
  • Shared folders and cloud applications
  • Remote-access tools
  • Administrative accounts

Remove or disable accounts that are no longer required. Confirm that current users have access appropriate to their roles. Document that the review was completed.

5. Check Whether Your PIAs Are Current

Compare your current operations with your existing Privacy Impact Assessments.

Have you changed software, vendors, locations, patient communication tools, information flows, business structure or Privacy Officer? Have you changed how health information is collected, used, disclosed, stored or destroyed?

Determine whether a PIA amendment, update or new PIA may be required.

6. Identify New Software, Vendors and Technology

Make a list of technology introduced since your last privacy review. This might include online booking, patient portals, electronic forms, payment systems, cloud storage, messaging tools, virtual care, transcription services or AI.

Privacy Officers should know which technologies handle health information and what safeguards are in place.

Review your Information Management Agreement listing and confirm that you have appropriate agreements with your service providers.

7. Find Out How Staff Are Using AI

Do not assume you know how AI is being used in your practice. Ask.

Staff and healthcare providers may be using AI for clinical documentation, meeting notes, drafting emails, summarizing documents, patient education, administration or marketing.

Confirm that identifying health information is not being entered into unapproved AI tools. Review whether your organization needs an AI Governance Framework, approved-tool list, vendor assessment, training or PIA.

8. Review Your Privacy Breach and Incident Log

Review the privacy incidents, near misses, complaints and concerns recorded by your organization.

Look for recurring problems, incomplete corrective actions, additional training needs and outstanding notifications.

For example, if faxes continue to be sent to the wrong recipients, reminding employees to “be more careful” may not be an effective corrective action. A recurring problem may indicate that the process or technology needs to change.

If you do not have a privacy breach log, now is a good time to start one.

9. Review and Test Your Privacy Breach Response

Review the Privacy Breach Management procedure in your Health Information Privacy and Security Policies and Procedures Manual.

Confirm that staff know how to recognize a possible breach, who to notify, what immediate containment steps to take and what information to document.

Test the procedure using a short, realistic scenario. A five-minute tabletop exercise can help your team understand what to do before a real incident occurs.

Document the exercise, who participated and any follow-up actions identified.

10. Document Your Review and Follow-Up Actions

For each checklist item:

  • Record what you reviewed
  • Identify any gaps
  • Assign responsibility
  • Establish a realistic follow-up date
  • Document when the action is completed

Keep the completed checklist as evidence of your ongoing privacy management program and demonstrable accountability.

Practical Privacy Tip

Do not try to fix everything during the check-up.

First, identify what is working, what has changed and what needs attention. Then prioritize your actions, assign responsibility and set realistic follow-up dates.

One completed action followed by another is more effective than an overwhelming list that never gets started.

Download the Privacy Officer Check-Up

Would you like a practical way to complete this review?

Download the Privacy Officer Check-Up: 10 Things You Should Review checklist.

Use it to recognize your accomplishments, identify gaps, assign responsibility and plan your next privacy management activities.

Get the checklist here:

Wish You Had a “Jeanie” to Help With Your Privacy Officer Role?

You don’t have to figure everything out on your own.

Practical Privacy Officer Strategies gives you practical training, tools, templates, and live coaching to help you know what to do next—and how to get it done.

Learn More About Practical Privacy Officer Strategies

How to Manage a Privacy Breach in Your Canadian Practice

How to Manage a Privacy Breach in Your Canadian Practice

​

How to Manage a Privacy Breach in Your Canadian Practice Workshop

Have you ever heard about a privacy breach at another practice and thought…

“I hope that never happens to us.”

The reality is — privacy breaches can happen in any healthcare practice, regardless of size, specialty, or technology. Whether it’s a misdirected fax, unauthorized access to a chart, lost device, or cyber incident, breaches are not a matter of if — but when.

What makes the difference is how prepared you are to respond.

I’m tickled pink to  partner with Kayla Das to deliver a live virtual workshop designed to provide practical, step-by-step guidance for Canadian healthcare practices. Kayla Das B.Rec, BSW, MSW, RSW is a trusted Business Coach For Therapists and Counsellors.

Live Virtual Workshop

How to Manage a Privacy Breach in Your Canadian Practice

In this interactive on-line session, we’ll walk you through what to do when a privacy breach occurs — before you ever have to face one in real time.

This workshop is ideal for:

  • Canadian clinic managers
  • Privacy officers
  • Practice owners
  • Social workers, counsellors, and mental health leaders
  • Clinical supervisors and consultants

If you are responsible for protecting patient information, this training will help you strengthen your breach response readiness.

What We’ll Cover

Participants will learn:

  • The difference — and overlap — between confidentiality and privacy
  • Legislative, regulatory, and professional practice requirements across Canada
  • Why privacy breaches are a significant risk you should prepare for
  • How to recognize when a breach has occurred
  • The 4-Step Response Plan for managing a privacy breach
  • Practical steps to prevent breaches before they happen
    … and more

 

Important for Ontario Practitioners

Ontario health information custodians (practice owners) are required to submit annual privacy breach statistics to the Ontario  Information and Privacy Commissioner by March 1 each year.

If you’re unsure what must be reported — or how to prepare — this workshop will address those requirements.

 Workshop Details

This on-demand mini-course, How to Manage a Privacy Breach in Your Canadian Practice is available now!
Mini-course is a replay from a live event hosted by Jean Eaton and Kayla Das on February 24, 2026 that is approximately 85 minutes.

Privacy breaches are stressful — but managing them doesn’t have to be overwhelming when you have a plan.

We hope you’ll join us for this practical, supportive session designed to help you protect your patients, your practice, and your professional reputation.

Texting with Patients: How to Do It Safely and Effectively

Texting with Patients: How to Do It Safely and Effectively

​Texting with Patients: How to Do It Safely and Effectively

Have you ever said…
“If only I had someone to ask!”

Each month, we tackle real questions from clinic managers, healthcare providers, and privacy officers inside Practice Management Success.

This month’s question:

Can you text your patients?

The short answer is:
Yes.

The better answer is:
Yes—but only if you do it thoughtfully, with the right safeguards in place.

Why This Matters

Texting is no longer “new.”

Patients expect it.
Staff rely on it.
And many EMRs now offer built-in messaging tools.

But here’s the problem:

👉 Texting is not always a secure communication method.

It’s difficult to:

  • Confirm who is sending or receiving the message
  • Control where the message is stored
  • Prevent miscommunication or disclosure

That means one quick message can turn into a privacy breach or medical error.

Start With Purpose (Not Technology)

Before you implement texting, ask:

Why do we want to use it?

Common reasons include:

  • Appointment reminders
  • Scheduling changes
  • Improving patient access
  • Reducing phone volume

These are all valid—but they are not all equal in risk.

From the Patient to the Clinic

Some clinics allow patients to text:

  • Appointment requests
  • Questions about care
  • Follow-ups

In some cases—especially remote or higher-risk populations—this may improve access to care.

But you must weigh this carefully.

👉 Sometimes the risk of not communicating is greater than the risk of using an unsecured method.

This is where your professional judgment—and policies—matter most.

What Are the Risks?

As the custodian, you assume the risk of using unsecured communication.

So your job is to:

  • Define acceptable use
  • Set clear boundaries
  • Train your team
  • Communicate expectations to patients

One of the most practical ways to do this?

👉 Create scenarios

  • When is texting appropriate?
  • When is it not?
  • What should staff do instead?

Document these decisions as part of your implementation plan.

Workflow Matters More Than You Think

If a patient texts your clinic—what happens next?

You need clear answers to:

  • Who receives the message?
  • On what device?
  • How is it verified?
  • How is it documented in the patient record?

If it’s not documented, it didn’t happen.

From the Clinic to the Patient

This is where most clinics start—and where risk is easier to manage.

Best use cases:

  • Appointment reminders
  • Basic instructions
  • Non-sensitive communication

Higher-risk uses:

  • Test results
  • Clinical advice
  • Sensitive health information

👉 Keep texting administrative, not clinical, unless you have a secure solution.

Consent and Patient Understanding

Patients must understand:

  • How texting works in your clinic
  • The risks to their privacy
  • Their role in protecting their information

This includes:

  • Using a personal phone (not shared or work devices)
  • Keeping their phone secure
  • Updating their contact information

Consent is not just a form—it’s a conversation and an agreement.

Use the Right Technology

Whenever possible:

  • Use EMR-integrated messaging
  • Avoid personal devices
  • Implement role-based access
  • Enable audit logs
  • Use multi-factor authentication (MFA)

These tools help you:

  • Maintain control of patient information
  • Improve workflow
  • Reduce manual documentation

Don’t Skip the PIA

Before you implement texting or email communication:

👉 Complete or update your Privacy Impact Assessment (PIA)

This doesn’t have to be overwhelming—but it is essential.

Your PIA should describe:

  • What you are implementing
  • How information flows
  • Risks and mitigation strategies
  • Policies and procedures

Practical Take-Aways

If you’re thinking about texting patients:

  • Start with low-risk uses (appointment reminders)
  • Use approved systems—not personal phones
  • Define clear rules and workflows
  • Train your team using real scenarios
  • Document everything
  • Review and adjust regularly

Want Help Getting Started?

If you want to go deeper, I’ve created tools to help you implement this safely:

✔ Sample texting authorization forms
✔ Step-by-step procedures
✔ Training resources for your team
✔ PIA guidance and templates

👉 Download the FREE report:
Can You Use Text Messaging with Patients?

👉 Get ongoing support:
Practice Management Success Membership

👉 Join me live:
Q&A with Jean –  2nd Tuesday of each month at 12 noon MT

Final Thought

Texting can absolutely improve access, efficiency, and patient satisfaction.

But it must be done with intention.

Because when it comes to privacy:

When we know better—we can do better.

Jean L. Eaton is constructively obsessive about privacy, confidentiality, and security especially when it comes to the handling of personal health information. If you would like to discuss how I can help your practice, just send me an email. I am here to help you.

Jean L. Eaton
Your Practical Privacy Coach
INFORMATION MANAGERS