As healthcare organizations race to realize the promise of generative AI, some find themselves stuck in a cycle of pilots and prototypes. While experimentation is critical, the landscape in 2025 and beyond will demand more than potential — it will require proof. Hospitals and health systems must shift focus from exploration to execution, embedding generative AI into clinical and operational workflows in ways that design and deliver measurable impact for clinicians and patients alike.
Whether health systems can keep up with the pace of innovation hinges on how decisively leaders can move from AI hype to practical value. In Wolters Kluwer’s Future Ready Healthcare Survey Report that examined generative AI and organizational priorities, 76% of healthcare professionals expressed a desire to reduce physician burnout — however, only 63% said their organizations are prepared to use generative AI to optimize departmental workflows.
To better understand what this transformation looks like in practice, Becker’s Healthcare spoke with Holly Urban, MD, MBA, vice president of business strategy at Wolters Kluwer Health.
Critical building blocks: thoughtful design + broader validation
Generative AI tools show great potential for improving healthcare. Innovative new tools have emerged — such as those that triage patient messages in EHR inboxes and reduce provider burden. However, without alignment to existing workflows and multidisciplinary involvement by stakeholders, even the most advanced solutions are likely to falter.
Dr. Urban explained that stakeholders must understand what workflows look like and who is responsible for what, especially when it comes to handoffs in care. For example, with a generative AI inbox tool, everyone needs to understand which messages will be routed to physicians and which will be routed to nurses.
Generative AI also has the potential to reduce the complexity of payer-provider relationships and lower the cost of healthcare. CMS, for instance, is examining how technology could be used to harmonize billing and coding workflows, especially in regard to prior authorization. Another area of opportunity is using generative AI to standardize payer contracts with providers.
According to Dr. Urban, there is a huge amount of disparity in these contracts that results in different rules. If contracts could be standardized, she said, it would reduce administrative complexity that’s costly for both providers and payers. That complexity represents healthcare dollars that are spent on administrative overhead, rather than contributing to patient care.
In this context, the success of any generative AI solution in healthcare requires upfront design and validation.
“Healthcare organizations are complex, interconnected ecosystems,” Dr. Urban said. “If you just drop GenAI on top of broken workflows, it will fall flat. The key to success is designing systems for the outcomes that you want to achieve. Teams and workflows must be aligned to support better collaboration and outcomes.”
Translating GenAI investments into real value: 3 best practices
To ensure generative AI initiatives drive operational value, Dr. Urban shared three best practices for healthcare organizations to consider.
- Establish strong governance. When hospitals and health systems have strong AI governance in place, they have greater confidence that the selection, piloting, implementation and monitoring of technology will be conducted in safe and ethical ways.
“Strong governance makes sure that patient safety concerns are accounted for, and it fosters clinician trust,” Dr. Urban said.
- Prioritize change management. While there is a lot of enthusiasm about generative AI solutions in healthcare, many physicians and other staff still don’t trust it. An intentional approach to change management can increase generative AI adoption rates by helping people understand the true purpose of the technology.
“Physicians always have to understand why a tool is important and how it will help them,” Dr. Urban explained. “Will it improve patient outcomes or enhance clinician efficiency so they can focus more on patients?”
Healthcare leaders can address silos by getting to the “why” and providing clinicians with data that show improved outcomes. Identifying local champions and “evangelists” can also help. Staff trust these individuals because they understand the workflows and nuances that are unique to particular healthcare organizations.
- Monitor performance continuously. The third way to generate operational value is by monitoring the performance of generative AI tools. Dr. Urban noted that with AI tools, organizations can’t just “set it and forget it.” Seeing clinical value from AI requires practical, real-world results. While the specific clinical, financial and operational outcomes that an organization is looking for are defined upfront as part of the governance process, teams must monitor those outcomes over time to determine whether generative AI tools are delivering the expected results.
“You need to monitor AI applications for bias, as well as overfitting the model. When an AI model learns on a training dataset that isn’t generalizable to different settings, the risk of hallucinations may increase,” Dr. Urban said.
There’s little doubt that generative AI can drive more efficient clinical workflows, reduce burnout, improve patient safety and more. But to turn this potential into lasting impact, healthcare organizations must move beyond experimentation. A structured deployment strategy — one that prioritizes governance, change management and continuous evaluation — will be essential to transforming generative AI from a promising innovation into a true force multiplier.Systemic strategies reinforce the value of collaboration and adaptable leadership. For further insights and actionable strategies to break down silos in healthcare, read Dr. Urban’s full article, “Systems thinking can help health leaders standardize clinical information for teams,” offering practical approaches for leaders to address these pivotal challenges.
