AI Momentum Grows in MedTech R&D: Leaders Share What’s Working and What’s Next

The June 26 R&D Leaders Circle meeting convened a cross-disciplinary panel of experts to explore how artificial intelligence is reshaping medical device innovation, and to launch the next round of Deloitte’s industry-leading benchmarking survey on R&D practices. Together, these sessions underscored the urgency and opportunity in aligning digital transformation with meaningful business outcomes across medtech.

The featured panel on AI in R&D brought together leaders from Google, NVIDIA, Level Ex, and Deloitte, as well as regulatory insights from a former FDA official. The conversation centered on how AI is beginning to deliver tangible value in medtech—not just in visionary applications, but in the day-to-day systems that underpin innovation.

Watch highlights from the Panel: “R&D in AI”

Panelists identified three major domains where AI is gaining traction in medtech development:

• Process productivity: AI is already proving useful in automating time-consuming, documentation-heavy processes—ranging from regulatory filing prep to supplier management and design change tracking. These applications are often the most accessible and offer the clearest ROI.

• Design productivity: Companies are using AI to assess manufacturability, simulate clinical workflows, and even evaluate risk in complex systems.

• Product intelligence: Applications like diagnostic algorithms, digital surgery, and intelligent physician training tools are on the rise but face more scrutiny due to safety and regulatory complexity.

Panelists emphasized that while the technology is ready, the main barrier to impact is cultural. Aaron Barthel, National Sales Manager at Google, described how teams still struggle with fragmented data and internal resistance. He noted that when organizations prioritize high-impact, low-risk AI use cases—like content generation and intelligent search—they begin to shift internal skepticism to momentum.

“If you use AI to generate data, you end up with bad results. But one of the things that we’ve gotten really good at is using math to generate synthetic data, which I do think is very promising.” – Sam Glassenberg, CEO, Level Ex CEO.

Sam Glassenberg built on this point, noting that many of medtech’s most persistent issues—like melanoma detectors or pulse oximeters that fail on certain skin tones—are fundamentally data problems exacerbated by technology gaps. “When you have a hammer, everything looks like a nail,” he said, warning against using AI on bad or biased datasets. Instead, he pointed to the gaming industry’s success with mathematically generating synthetic data as a way forward. Techniques once used to simulate lifelike environments in video games are now helping to train both clinicians and AI models in healthcare—especially in areas where diverse or complete real-world datasets are lacking. Brendan O’Leary, formerly with FDA, reinforced this approach, stressing that confidence is built through repeated success in controlled applications. “Start where you can win,” he advised, echoing a theme heard throughout the session.

Renne Yao of NVIDIA reflected on her experience working with thousands of health AI startups, several of which have shortened regulatory timelines from years to months by leveraging digital tools strategically. She noted a growing divide: organizations that successfully integrate AI into their operations tend to do so through lean, nimble teams with clear mandates, executive support, and an openness to redefining their business models. Sam Glassenberg, CEO of Level Ex and a former game developer, spoke passionately about the value of giving innovators the freedom to push boundaries. “The people who say ‘it feels like cheating’ to use AI for hard problems,” he said, “are the ones who’ll get left behind.”

“The change management—bringing people along—is actually the biggest battle. It’s not the technology itself. Training on where AI can truly add value and really address the fear in human beings is going to be critical in this next era of innovation.” -Renee Yao, Global Healthcare AI Startups Lead, NVIDIA

While the enthusiasm for AI was high, panelists were quick to acknowledge that adoption doesn’t mean indiscriminate use. “Knowing where AI will fail is just as important as knowing where it will succeed,” said O’Leary. Trust, reliability, and transparency—especially in regulated environments—must be built deliberately, and organizations need to establish internal policies that allow for experimentation without compromising compliance or safety.

Ultimately, successful AI adoption will depend not only on choosing the right problems to solve, but also on equipping teams with the tools, trust, and leadership support needed to scale those solutions beyond isolated pilots.

Following the panel, Doug Billings, Managing Director at Deloitte, presented the 2025 relaunch of their medtech-focused R&D benchmarking survey. Now in its third round, the study is designed to help business units compare their performance across a wide range of innovation and product development metrics.

The study goes beyond high-level R&D spending data to examine product development outcomes by cohort (e.g., disposables, implantables), cycle times, predictability, digital practice maturity, and functional staffing ratios. New additions this year include geographic analysis of R&D hubs and expanded coverage of digital business models—recognizing the growing importance of software and service-enabled offerings.

Of note, participants receive confidential, custom analyses that compare their results to high-performing peer groups. Billings emphasized that participation is confidential and free, with the only investment being time and effort to complete the survey workbook. He invited all present to take part in the study and spread the word across their networks, underscoring that greater participation will lead to more robust insights for the entire community.

The survey will remain open for the remainder of the year, with the possibility of previewing early trends at MDIC’s November events. For those not ready to commit immediately, Deloitte will follow up regularly to support onboarding and participation.

The meeting left no doubt that artificial intelligence is becoming inseparable from R&D in medtech. But the most actionable insight was also the simplest: start small, but start now. Successful organizations are focusing on practical, achievable use cases—especially in process automation—where early wins build momentum. Teams that pair AI with engineering expertise are not just making their workflows more efficient; they’re solving hard problems that previously felt out of reach.

At the same time, initiatives like the Deloitte benchmarking survey give leaders a chance to step back and ask critical questions about how their organization compares to the broader field—what’s working, what’s lagging, and what’s next.

Whether through pilots in content generation, investments in digital simulation, or contributions to industry-wide benchmarks, the message was clear: the future of medtech innovation depends on leadership willing to experiment, iterate, and share.

Interested in joining the R&D Leaders Circle? Learn more and fill out our interest form for an invite to get involved.