“AI-powered” once signaled that a healthcare company was doing something novel. Today, it often tells us almost nothing.
The term now appears across nearly every corner of healthcare, from diagnostic platforms and clinical decision-support tools to patient engagement apps, mental health services and administrative software. Yet two products described as “AI-powered” may have little else in common. However, what they are programmed to do determines what evidence is needed, what risks must be addressed and what patients and healthcare professionals should reasonably expect from the technology.
This issue is receiving greater attention throughout the healthcare community. The American Academy of Nursing’s 2026 Health Policy Conference, held earlier this month, focused on the promise and responsibility of digital health, including questions of ethics, accountability and the patient relationship. On October 22 and 23, the HIMSS AI in Healthcare Forum will continue the conversation with a focus on governance, policy and real-world implementation.
For healthcare communicators, calling something “AI-powered” is no longer enough. If companies want patients, clinicians, investors and other stakeholders to understand the value of their technology, they need to explain what the AI does and why its role matters.
Start with the task, not the technology
The first question should be the simplest: What does the AI do?
A vague description may sound sophisticated, but it usually makes the technology more difficult to understand. Saying that a platform “uses advanced AI to transform patient care,” for example, leaves nearly every meaningful question unanswered.
A stronger explanation identifies the specific problem, the task performed and the intended user. Descriptions give audiences something concrete to evaluate. They also help prevent companies from overstating the role of AI. In many products, AI is one component of a much larger system involving clinicians, researchers, care coordinators, patients and established workflows. Describing that system accurately can be more persuasive than positioning the algorithm as the entire story.
Define what “works” means
Once the task is clear, the next question is whether there is evidence that the technology performs it successfully.
The right evidence will depend on the claim. Healthcare companies should be precise about the difference between what has been demonstrated and what is still being studied. Was the technology evaluated internally or independently? Was it tested in a controlled environment or incorporated into routine care? How many people or sites were included? What comparison was used? Are the findings preliminary, peer-reviewed or based on real-world experience?
Not every communications asset needs to answer all of these questions. But the company should be prepared to answer them, and its public claims should reflect the maturity of the evidence.
Words such as “may,” “designed to” and “demonstrated” are not interchangeable. Neither are “accurate,” “effective” and “clinically meaningful.” In healthcare communications, those distinctions protect credibility.
Explain what changes for people
Even when an AI tool performs well technically, audiences still need to understand what difference it makes. Companies often spend so much time explaining the sophistication of the technology that they forget to establish how the experience of care changes because of it.
That question is especially important in mental health. An AI-enabled mental health tool may provide education, help someone reflect on how they are feeling or connect them with additional support. But its value cannot be communicated responsibly without explaining its intended role. Is it a wellness resource, a peer-support tool or part of a clinical service? What happens when someone expresses signs of a crisis? When does a human become involved?
The same principle applies to diagnostics. Faster analysis may sound beneficial, but speed alone is not necessarily the outcome that matters. The more important question may be whether the technology helps the right patient reach the next step in care sooner and whether the healthcare system has the capacity to provide that care.
Be direct about limitations
Credibility does not require claiming that a technology is infallible. In fact, acknowledging limitations can make an AI story more trustworthy.
Healthcare companies should be able to explain where a tool has been tested, where it has not, what types of errors can occur and what safeguards exist when the system is uncertain or wrong. They should also make clear whether performance varies across populations and how the technology is monitored after implementation.
The World Health Organization’s guidance on AI in health emphasizes transparency, safety, accountability, autonomy and equity. These principles are not only regulatory or ethical considerations. They are communications considerations as well. People are more likely to trust a company that explains where its technology fits than one that presents AI as magic.
Replace the label with a defensible story
Healthcare companies do not need to stop talking about AI, they just need to talk about it with greater specificity.
Before calling a product “AI-powered,” communications teams should be able to answer six questions:
- What specific task does the AI perform?
- Who uses it, and where does it fit into the healthcare workflow?
- What evidence supports the claims being made?
- What changes for patients, healthcare professionals or health systems?
- What are the technology’s limitations and safeguards?
- Who remains accountable for the decisions and outcomes associated with its use?
These questions should be addressed early, with input from clinical, technical, regulatory and communications teams.
As the use of AI technology becomes more common, the AI label itself will carry less weight. The companies that stand out will move beyond “AI-powered” and tell a more useful story: what their technology does, how they know it works and why it meaningfully improves the experience of care.
