Podcast Summary: Becker’s Healthcare Podcast with Nabile M. Safdar, Chief AI Officer, Emory Healthcare
Episode Date: January 8, 2026
Host: Grace Lynn Keller
Guest: Dr. Nabile M. Safdar
Episode Overview
This episode features Dr. Nabile M. Safdar, Chief AI Officer at Emory Healthcare, in conversation with Grace Lynn Keller, live from the 10th Annual Health IT, Digital Health, and RCM Meeting. The discussion explores the rapidly evolving landscape of artificial intelligence in healthcare, including practical use cases, challenges in governance, the regulatory environment, and top advice for industry leaders as they navigate future technological advancements.
Dr. Safdar’s Background and Role
[00:49]
- Radiology and Informatics Roots:
Dr. Safdar details his background as a radiologist with a strong interest in imaging and clinical informatics, disciplines at the forefront of digital transformation in medicine. - Transition to AI Leadership:
“Radiology is one of the most digital specialties ... many people have speculated that we could go away as a profession. Well, I was interested in that out of self preservation and I thought I'd learn more.” (Safdar, 01:04) - Responsibilities:
- Leads AI system governance, implementation, deployment, and evaluation at Emory
- Manages AI efforts across clinical, operational, marketing, and patient/internal-facing applications
Key Use Cases of AI at Emory Healthcare
[02:15]
- Ambient AI & FDA-Cleared Tools:
- Success in adopting FDA-cleared AI technologies in cardiology and radiology, with an eye towards pathology
- Agentic Chatbots for Patient Engagement:
- Early positive results using chatbots for:
- Prescription reminders post-discharge
- Home blood pressure monitoring and escalation protocols
- Patient outreach, providing guided next steps when readings are abnormal
- “We see early success with agentic chatbots ... reaching out to patients to help remind them ... and making sure that they're in a range which is safe for them and if not, quickly escalating to a human...” (Safdar, 02:58)
- Early positive results using chatbots for:
- Wider Adoption of LLMs:
- Noted demand from staff for tools like ChatGPT, Claude, and Gemini both personally and professionally
- Emory implementing “HIPAA safe harbor tools” to safely extend LLM efficiency to healthcare workflows
- Emerging Frontiers:
- Early exploration into revenue cycle, marketing, and access using AI
Governance Strategies: Balancing Innovation and Risk
[05:06]
- Risks of Extreme Approaches:
- Lack of governance can lead to institutional risks
“One is the absence of AI governance ... that looks like basically everything goes ... data deals at will, AI deployments at will.” (Safdar, 05:13) - Overly restrictive governance stalls progress
“Governance can be so restrictive that a single committee becomes a choke point ... Any large system needs to be more nimble than that.” (Safdar, 05:33)
- Lack of governance can lead to institutional risks
- Emory’s Model:
- Decentralized, department-level AI governance aligned with institutional priorities
- Empowering functional experts with guidelines for local decision-making
- Escalation protocols for higher-risk or more impactful AI deployments (patient-facing, highly autonomous, conflict-of-interest concerns)
Navigating the Evolving Regulatory Landscape
[07:25]
- Complex Web of Federal and State Oversight:
- Agencies involved: Congress, state legislatures, CMS, HHS, ONC, OCR
- Over 200 state legislative initiatives concerning AI noted
- Recent Key Legislation:
- Tracking California’s SB53: Requires large AI model developers to release safety protocols
- Evolving FDA regulatory stance on healthcare AI
- Advice to Leaders:
- Proactive monitoring of regulatory developments
- Close collaboration with regulatory offices and vendor partners
- “My advice is to keep an eye on it, to talk to your vendor partners, to talk, if you have an office of regulatory affairs, to have close coordination ... and potentially even impact your own state's legislation, if appropriate.” (Safdar, 09:23)
Top Advice for Healthcare Leaders Adapting to Technological Change
[10:02]
- Agility is Crucial:
- “It is harder and harder to look out the front windshield and see far down the road. Things are changing fast. So it's a curvy road and you need to be very agile and nimble.” (Safdar, 10:05)
- Alignment Across Leadership:
- Integrate digital and AI governance into broader corporate governance strategies
- Front-Line Collaboration:
- Engage nurses, physicians, and administrators early in AI solution development
- Emphasizes co-design and integrated workflows to ensure adoption and measurable success
- “...they're a part of the process, they're engaged early on, they're co developing solutions, co creating solutions. Those are integrated well into their workflows and then we celebrate the wins...” (Safdar, 10:49)
Notable Quotes
-
On Digital Transformation in Radiology:
“Radiology is one of the most digital specialties and one of the areas which has been early adoption … many people have speculated that we could go away as a profession.” — Nabile M. Safdar [01:04] -
On Governance Balance:
“Any large system needs to be more nimble than that. So what we've had success with is helping specific functional areas ... to develop their own AI governance.” — Nabile M. Safdar [05:39] -
On Speed of Change:
“It is harder and harder to look out the front windshield and see far down the road. Things are changing fast. So it's a curvy road and you need to be very agile and nimble.” — Nabile M. Safdar [10:05]
Timestamps for Key Segments
- 00:49: Dr. Safdar’s background and journey to AI leadership
- 02:33: Most impactful current AI use cases at Emory
- 05:06: Strategies for navigating AI governance and sharing organizational examples
- 07:25: Impact of recent state and federal regulations; Emory’s regulatory approach
- 10:02: Advice to healthcare leaders preparing for ongoing change
This episode is an insightful primer for healthcare executives and innovators navigating AI adoption, highlighting practical applications, real-world governance models, regulatory complexities, and the human-centric approach required for successful digital transformation.
