PITTSBURGH -- The nation's top health information technology office is working to lay the groundwork for a more connected, AI-ready healthcare system. But the road ahead is complicated, according to a federal official who spoke at the Society for Imaging Informatics in Medicine (SIIM) annual meeting.
Jason Funderburk, MD, deputy national coordinator for health IT at the Office of the National Coordinator for Health Information Technology (ONC), joined SIIM incoming board chair Alex Tobin, MD, for a wide-ranging fireside chat that touched on image sharing barriers, information blocking, AI governance, provider burnout, and the future of health technology policy.
Jason Funderburk, MD (right), shares his thoughts on image sharing, AI, and what emerging technologies will bring in the near future in a fireside chat with incoming SIIM board chair Alex Towbin, MD (left).
A practicing interventional radiologist with more than 20 years of clinical and leadership experience, including a stint as chief of staff for the Salt Lake City Veterans Affairs (VA) Healthcare System in Utah, Funderburk brought a grounded perspective to topics that often remain abstract in policy circles.
"Our job is to make policy changes that actually have real impact on physicians," he said. "Being a physician, we now have at least three physicians at the table in policy discussions, which is a change. We have a voice pushing things forward for patients."
Ditching the disk
One of the session's exchanges centered on medical image sharing, a challenge the imaging informatics community has been trying to solve for about two decades.
"We've been trying to ditch the disk for 20 years and we still have disks," Tobin said. "And maybe even worse, we're bringing more disks into other specialties as we try to standardize."
Funderburk acknowledged the frustration, attributing the persistent problem to a mix of infrastructure gaps, proprietary data concerns, liability ambiguity, and situations where digital sharing is technically workable but isn't happening.
He noted that ONC does not currently have direct regulatory authority over imaging systems like PACS, and that questions about how to extend its reach into diagnostic imaging are actively being discussed.
"I think there is enough interest and enough belief that this is worth addressing for utilization, for safety, and for cost," he said. "The right people are at the table now to have this conversation."
For specialties like dermatology and pathology, where no universal imaging standard equivalent to DICOM exists, the problem is even more acute. And even where DICOM is in use, as in ophthalmology, Funderburk acknowledged that compliance in format doesn't always translate to useful, actionable data.
"Now you see the trickiness of policy," he said.
Information blocking and the trust problem
Funderburk also addressed information blocking, one of ONC's areas of regulatory enforcement, describing it as a nuanced issue that is often misunderstood by well-intentioned organizations.
He noted that for the first time in ONC's history, the agency recently began sending non-conforming letters to organizations flagged for potential information blocking, signaling a more active enforcement posture.
On the question of patient results being released to portals before ordering physicians have been notified, Funderburk acknowledged the tension between patient autonomy and the practical realities of care coordination. This is particular in sensitive cases involving adolescents, where privacy rights and parental access can conflict.
"It's challenging," he said. "I wish there was a perfect solution right now."
AI: Make the data accessible
When the conversation turned to AI, Funderburk described ONC's role as building the infrastructure that allows AI to function rather than regulating the tools themselves, which falls under the U.S. Food and Drug Administration (FDA).
"The biggest conversation we're hearing is just make sure the data is accessible for AI," he said. "If you want to deploy a product, you need the ability to tie into that data wherever it is."
And with national debt levels straining federal budgets, Funderburk said the pressure is on technology developers and policymakers to show that AI can reduce healthcare costs and not just increase use.
"Let's deploy technology in a way that actually decreases costs, increases quality, and improves provider satisfaction," he said. "There are tools out there right now that, if we listen to providers as they're integrated into workflows, can actually make the day better."
Looking five years out
Asked where health IT will stand in five years if current efforts succeed, Funderburk offered a vision centered on prevention, provider relief, and truly patient-centered data access.
"Let's not waste this," he said. "This is such an amazing opportunity. Let's focus on the right things [like] patient outcomes, prevention for once, and letting providers get back to taking care of patients."
He envisions a future where personal health data flows seamlessly to individuals' devices, enabling earlier interventions and more meaningful conversations with their care teams. This is also where radiologists evolve into stewards of AI-assisted information rather than being displaced by it, he added.
"Radiologists are profoundly adaptable," Funderburk said. "I think it'll be a good journey for us, ultimately."
ONC is holding a webinar on June 25 to review findings from its recent AI Request for Information. More information can be found on ONC’s site.
Check out AuntMinnie’s full coverage of SIIM 2026 here.
