13398
Blog
Andy Pollen

A Candid Conversation About Clinical AI Adoption, Governance and Real-World Impact

C-Suite, Customer Success, Information Technology
Table of Contents

For many health systems, the challenge isn’t just whether to adopt clinical AI, it’s how to do it effectively. 

In a recent webinar moderated by Larry Vernaglia, Partner and Health Care Lawyer from Foley & Lardner LLP, leaders from Ochsner Health, the Coalition for Health AI (CHAI) and Aidoc addressed that central question: How can health systems implement AI at scale — responsibly, efficiently and sustainably?

Below is a summary of key themes and takeaways, but you can also access the full webinar.

AI Policy Is Taking Shape — But Not Where You Might Expect

Brenton Hill, Head of Operations and General Counsel at CHAI, opened with a policy update: The federal government is taking a cautious approach to AI regulation. Despite a few executive orders and rulemaking efforts (like updates to Section 1557 of the Affordable Care Act (ACA), enforcement is likely to be minimal in the near term. Instead, the regulatory action is unfolding at the state level, with California, Colorado and Texas leading the way.

Bottom line: Health systems and vendors should prepare for a patchwork of state-level rules, likely using the most stringent state laws as the baseline for compliance.

Choosing AI Tools: Value, Transparency and Workflow Fit Come First

Ashley Weber, Vice President of Ancillary Services at Ochsner Health, emphasized the importance of choosing solutions that solve real problems and that integrate cleanly into existing workflows. “I’ve seen at least 15 very relevant AI applications come across my desk…But we have finite resources. If I bought all 15, would we see the same ROI — financially or in quality? Could we even support them? They’re often time-consuming to implement…and that’s before we even get to adoption.”

Key selection criteria shared by all panelists:

  • Transparency: Vendors should provide clear documentation, including model performance, data sources and potential risks.
  • Workflow integration: If a tool disrupts existing processes or adds cognitive burden, adoption will suffer.
  • Demonstrable outcomes: The ability to show retrospective performance on your own data.
  • Scalability: Start with a few high-value use cases, then expand from there.

Successful Implementation Requires Local Customization and Change Management

There’s no such thing as a one-size-fits-all rollout. As Weber noted, “Healthcare is local.” Even in a large system like Ochsner, workflows vary by site, staffing mix and patient population.

Aidoc’s Chief Technology Officer Demetri Giannikopoulos echoed that view, stressing the importance of adaptable platforms over rigid point solutions: “The same pulmonary embolism (PE) workflow at Yale might not work at Mount Sinai. You need to have the ability to deliver a platform and a solution that’s flexible and able to adapt inside the environment.” 

Implementation must address:

  • Local workflow realities
  • Multi-specialty coordination
  • Clinician education and buy-in
  • Communication guardrails and legal considerations 

Governance Must Be Tiered, Risk-Informed and Continuous

Some best practices shared by the panelists included: 

  • A formal AI policy defining roles, responsibilities and accountability
  • A tiered risk approach (not every algorithm needs the same level of oversight)
  • Ongoing performance monitoring, bias detection and incident reporting
  • Active data governance: who owns the data, how it’s used and how outputs are handled
  • Vendor accountability baked into contracting and business associate agreements

Weber noted that Ochsner uses a dedicated AI Center of Excellence and governance committee, closely aligned with its broader IT and data governance structures. Even so, she emphasized that this structure is evolving — and that smaller hospitals may need to take a leaner approach.

Clinician Adoption Hinges on Trust and Culture

To gain real traction, AI tools must be trusted by the people using them. That means:

  • Performance must meet or exceed clinical standards
  • Workflow fit must be near-seamless
  • Leaders must create a dialogue with staff to understand their concerns and motivations

Weber called this “a culture of technology” — one rooted in clarity of purpose and aligned with the organization’s mission to serve patients. Hill added that clinician engagement begins long before deployment. “You need to walk the halls…” to understand their pain points and build trust over time.

Sustaining AI Requires Ongoing Monitoring — and Shared Responsibility

Post-implementation, AI performance must be measured and maintained.

Ochsner holds monthly reviews to evaluate algorithm performance, drift and value — aligning outcome tracking with existing clinical KPIs. Contracts with vendors are expected to include shared accountability for ongoing monitoring and bias detection.

Giannikopoulos summed it up: “It’s our job as vendors to help health systems stay ahead.”

Final Reflections: The Future of AI in Healthcare

What comes next? Panelists offered a mix of pragmatic and visionary views:

  • Agentic AI (autonomous agents) will change how work gets done — especially in admin-heavy functions, like the revenue cycle.
  • AI won’t replace clinicians but will force reinvention of how care is delivered and how roles are structured.
  • The ethics of automation will become increasingly important, particularly as certain support roles could be impacted.

As Weber put it, “We don’t see AI as a replacement. We see it as a way to ease the very real stress our people are feeling today.”

Want to hear more insights from the panelists? View the on-demand webinar.

Explore the Latest AI Insights, Trends and Research

Follow a manual added link

Webinars

Aidoc experts, customers and industry leaders share the latest in AI benefits and adoption.

Follow a manual added link

eBooks

Explore how clinical AI can transform your health system with insights rooted in real-world experiences.

Follow a manual added link

AI Strategy

Learn how to go beyond the algorithm to develop a scalable AI strategy and implementation plan.

Follow a manual added link

Solutions

Explore how Aidoc can help increase hospital efficiency, improve outcomes and demonstrate ROI.

Andy Pollen
Andy Pollen is an experienced healthcare communicator and strategist who currently serves as the Director of Marketing Communications for Aidoc. Previously, he was the global marketing communications lead for critical care solutions within 3M Health Care's Medical Solutions Division, now Solventum. Pollen has also held communications positions with the University of Minnesota Academic Health Center, Indiana University Health and several business functions within Eli Lilly and Company through Borshoff, a creative services agency. He earned a bachelor’s degree in public relations and journalism from Ball State University and holds a master’s degree in business administration from Anderson University.
Andy Pollen
Director, Marketing Communications

© Copyright 2025 Aidoc