Making AI-Powered Healthcare Communication Safer for You and Your Family.

MedStar Health Research Institute led a national collaborative to craft the AI Care Standard for Patient Communication, a new framework designed to help healthcare systems evaluate and safely adopt patient-facing artificial intelligence tools. 

I spend my professional life studying how technology, healthcare, and humans interact. But sometimes I’m a patient too! So when I recently got some test results, I turned to artificial intelligence to help me understand what it meant. 

The tool was helpful, and like many people, I got a simplified explanation of complex clinical information. But most people don’t have a background in research or healthcare. It would be easy to overlook errors or generalities in the AI output or misunderstand its message. 

As more AI healthcare tools become available to the public, MedStar Health Research Institute is working to make them safer and more helpful for patients.

I’m proud to be a co-chair of the PatientAI Collaborative™, a voluntary, non-commercial coalition of doctors, technology experts, and patient advocates. We’ve launched the AI Care Standard™ for Patient Communication, a first-of-its-kind guide to help ensure that patient-facing AI tools provide accurate, clear information.

What is patient-facing AI?

Patient-facing AI isn’t behind-the-scenes software that reads imaging reports or transcribes notes. Instead, it’s digital tools that communicate with you and your family directly, such as:

• Interactive chatbots, like those found in mental health apps or on clinic websites

• Automated text messages or phone call reminders

• Draft replies to some messages sent through your patient portal

Healthcare AI is currently growing faster than safety policies at the federal level can keep up. State rules governing AI are inconsistent. They can be very different from state to state. When digital tools are put in place without good safety rules, the AI tool could:

• Share incorrect or harmful mental health guidance or medical advice 

• Fail to inform you that you’re talking to a machine

• Fail to notice when you’re describing a true medical emergency

Patient-facing AI tools have potential to be very useful in managing your health. But to offer these tools safely, guidelines are critical. That’s where the AI Care Standard comes in.

Related: Read “Why Your Doctor Is Better Than an AI Search for Health Advice.”

What you should expect when you interact with healthcare AI.

Whether you’re using an AI-powered chatbot, an automated text messaging service, or a patient portal tool, you should always feel confident the technology is working in your best interest. You should be able to trust that it:

• Recognizes emergencies: The tool instantly stops an automated conversation and connects you to a real nurse or doctor if you describe severe symptoms or seem distressed

• Tells you when you are interacting with AI: The technology clearly states that you are talking to a computer, shares where its health facts come from, and openly admits when it does not know an answer

• Explains information clearly: AI avoids confusing medical language, uses simple everyday words, and checks to make sure you understand your care instructions

• Encourages you to contact your care team: The tool explains your options but always reminds you to speak with your personal doctor before making any final health decision.

These expectations form the foundation of the new AI Care Standard. By building these guardrails, we can give healthcare systems across the country a shared model for ensuring that any AI tool interacting with you meets these essential safety measures. 

You shouldn’t have to wonder if an AI tool is safe.

You should never have to wonder whether an AI tool is giving you accurate health information, nor if it’s been properly tested before being used in your care.

To give you peace of mind, the AI Care Standard includes an interactive toolkit called the Evaluation Framework. This tool means hospitals don’t have to guess or take a company’s word that a product is safe. Instead, it gives teams a step-by-step method to audit an AI system before it ever interacts with a patient.

This rigorous testing process makes sure that before a digital tool reaches your screen, your healthcare team knows:

• It was built for everyone: The tool was trained and tested using a truly diverse patient population

• It knows its limits: The software clearly admits when it does not know an answer or lacks confidence in a response

• It stays in its lane: The tool never tries to answer complex clinical questions that should be out of bounds

• It has proper oversight: Real human clinicians are built into the process to oversee how the technology functions

By evaluating tools carefully behind the scenes, healthcare organizations can give you confidence that any AI communication tool your doctor uses has earned its place in your care. 

Related: Read “How Health Systems and Policymakers Can Prioritize Patient Safety When Integrating AI.”

Why safety should be built-in from the start.

It’s easy to assume that as long as a doctor or nurse double-checks an AI generated message before it reaches you, everything will be fine. But at the National Center for Human Factors in Healthcare, where we study how humans and technology work together, we know that relying just on a double-check isn’t a true safety net.

When you contact your doctor’s office, you want your care team focused entirely on you—your diagnosis, your questions, and your emotional well-being. Asking clinicians to spend hours proofreading computer-generated text takes their eyes off what matters most: your care. 

Safety can’t be a last-minute chore added at the end of the line. By building strict standards into AI tools from the start, we ensure the technology is safe. This allows your healthcare team to give you the personal, human attention you deserve.

Related: Read “Research Explores the Value of Rapid Testing of AI in Healthcare.”

What this means for you and your family.

At MedStar Health, our priority is making sure healthcare feels easier and less overwhelming for you, not more complicated or impersonal. 

We know that everyone feels different about technology. For instance, some people still hesitate to put credit card information online, while others are hyper-connected to texting apps. Whether you prefer a face-to-face conversation with your doctor or enjoy the convenience of a digital tool, we respect your preferences.

When used correctly and safely, AI has incredible power to make your life simpler. By establishing the AI Care Standard, we’re making sure any digital helpers you encounter at MedStar Health have passed strict safety checks. As technology grows smarter, our promise remains unchanged: keeping your care safe, clear, and deeply human.