Navigating the New MedTech Playbook: The Expectation of AI in MedTech

Navigating the New MedTech Playbook is a series from Stifel Venture Banking’s Life Sciences & Healthcare team focused on the operational and financial realities shaping today’s medtech market. Drawing on conference takeaways, client work, investor conversations, and patterns observed across the life sciences ecosystem, the series explores the strategies, decisions, and market shifts shaping how medtech companies scale in a more disciplined capital environment.

Medtech, traditionally a sector defined by hardware innovation, has not been immune to the transformative impact of AI. Over the last few years, the software layer, or “AI-enabled,” aspect of medtech has been increasingly garnering attention, driving meaningful advances across areas such as imaging, screening, monitoring, and robotics. However, as the market has matured, so has the framework for evaluating these technologies.

A few years ago, AI itself was a differentiator in MedTech investment conversations. Founders who could credibly describe an AI-enabled capability had a meaningful edge in investor discussions. 

That era is over. Today, investors are inundated with decks showcasing AI-enabled companies, and businesses that are not incorporating AI into their products, workflows or processes have become increasingly rare.

As a result, the questions investors are asking have shifted fundamentally. They don’t want to know if a company uses AI. They want to know if the company can demonstrate what its AI actually does, what makes their data uniquely positioned to deliver new capabilities, and how they are able to validate it in real clinical settings, with measurable outcomes tied to reimbursement logic.

Companies that can demonstrate meaningful results, differentiated data sets, and clear evidence of value creation are attracting outsized investor attention. Companies that can’t are facing a level of scrutiny that the “we use AI” narrative no longer deflects.

Understanding where the bar has moved, and why, is increasingly essential for founders building AI-enabled MedTech companies and for investors evaluating them.

The scale of what’s already been built

The pace of AI innovation in MedTech continues to accelerate. According to Innolitics’ 2025 Year in Review, the FDA authorized a record 295 AI-enabled medical devices in 2025, bringing the cumulative total to 1,451 authorized devices. Radiology accounted for 71.5% of all authorized devices, with cardiovascular and neurology representing the next-largest clinical specialties.

The vast majority of AI-enabled devices are being authorized through 510(k) pathways, allowing companies to demonstrate substantial equivalence to predicate devices, with De Novo and PMA pathways accounting for a much smaller portion of authorizations. However, clearance alone is increasingly viewed as insufficient. Investors, providers, and health systems are placing greater emphasis on clinical validation, real-world performance, and measurable outcomes beyond what is required for regulatory clearance, knowing those are the factors necessary to achieve successful reimbursement in the near/mid term.

On the capital formation side, venture capital funding trends continue to highlight growing investor interest in AI-enabled healthcare companies. According to Bessemer Venture Partners’ State of Health AI  2026 report, AI companies captured 55% of all health tech funding, up from 37% in 2024, 33% in 2023, and 29% in 2022. The direction of travel is clear: an increasing share of healthcare venture capital is being directed toward companies with AI-driven products, workflows, and business models.  

On the other end of the spectrum, recent medtech acquisitions also suggest growing strategic interest in AI-enabled capabilities, particularly in imaging, decision support and robotics. Recent transactions such as Medtronic’s acquisition of Cathworks ($585M, April’26), GE Healthcare’s acquisition of Icometrix ($98M, Nov’25), and Zimmer Biomet’s acquisition of Monogram technologies ($378M, Oct’25) illustrate interest in AI-enabled capabilities spanning cardiovascular disease management, neuroimaging and orthopedic robotics.

Given how quickly AI is evolving, where we are as an industry is likely to be drastically different even 12-24 months from now. Capabilities that seem aspirational today may become standard practice in the near future. Today’s AI helps cardiologists interpret CT, MRI and EKG data, while tomorrow’s AI may help cardiologists routinely evaluate treatment pathways on personalized digital cardiovascular models before making decisions in the real world. While predicting winners remains difficult, we expect AI-enabled technologies to continue to draw a growing share of investment and M&A interest.

What investors are evaluating

The investor lens on AI in MedTech has sharpened considerably over the past 18 months. Four criteria show up consistently in the diligence conversations that matter most.

Aptitude for adaptability

A key focus for investors across industries and market cycles is the management team. Time and again, strong teams have outperformed stronger products, while weaker teams have failed to capitalize on compelling technologies. However, AI and the rapidly changing AI environment are presenting an entirely new set of challenges and considerations for founders to navigate.

The strongest management teams are assembling organizations that combine deep technical expertise, clinical understanding, commercial execution, and an engrained willingness to continuously adapt.  In a market where capabilities evolve so rapidly and competitive advantages can compress quickly, learning velocity and execution arguably matter as much as the underlying technology itself.

Questa points to its portfolio company Avive as an example of this philosophy in practice. Rather than simply incorporating AI into an existing product, the company’s highly technical team developed its AED platform from the ground up, applying machine learning to arrhythmia classification to enable faster, versatile, and more accurate rhythm diagnosis.

As the time between first-generation and subsequent generation products continues to shrink, the companies that have the right teams in place to rapidly iterate, adapt, and improve without sacrificing quality will have a clear competitive advantage. Increasingly, investors are betting not only on the product itself but also on the team’s ability to build future versions of that product.

Validated clinical impact

Another important question investors are asking about an AI capability is whether its impact has been validated in real clinical settings. Not in a controlled research environment. Nor solely in a company-sponsored study with a limited sample. In the settings where the product will actually be used, with the patient populations it will actually serve, producing the outcomes it claims to produce.

This standard has raised the bar significantly for early-stage AI-enabled companies. A 2025 JAMA Network Open study of 903 FDA-approved AI devices found that clinical performance studies were reported for only about 56% at the time of approval, with roughly 24% explicitly stating none had been conducted.

That bar may continue to evolve. Notably, a recent FDA discussion paper suggests that generative AI medical devices may require a different evaluation framework than traditional software. The agency emphasizes the importance of ongoing postmarket performance monitoring and is exploring competency-based approaches that would assess not only accuracy but also safety behaviors, clinical proficiency, robustness, and real-world performance throughout a device’s lifecycle.

Investors are increasingly aware of these considerations and are asking harder questions about the evidence base behind AI claims, particularly for companies that haven’t yet navigated the full regulatory pathway.

Questa points to its own portfolio for proof. Eko Health and EnsoData were each built AI native, and each has converted that orientation into an accelerating cadence of validated product milestones. Eko has secured successive FDA clearances for a suite of cardiac AI algorithms since its first 2020 clearance, while EnsoData has expanded a single sleep scoring clearance into multiple spanning the sleep care pathway.

The practical implication for founders is that “our AI improves diagnostic accuracy” is not a sufficient claim. “Our AI reduces false negative rates in lung nodule detection by 31% in a peer-reviewed study of 2,400 patients at three academic medical centers” is a different conversation entirely.

Workflow integration

An equally important consideration is whether the AI fits into how clinicians work. 

FDA clearance is a necessary but insufficient condition for adoption. The history of MedTech is filled with clinically effective products that struggled commercially because they added friction to clinical workflows rather than reducing it.

Investors evaluating AI-enabled MedTech companies are spending increasing time on questions like: 

• How does the product actually get used in a health system? 
• What does the integration with EHR and PACS systems look like? 
• Who in the health system makes the adoption decision, and what is the procurement process? 
• How long does implementation take, and what does ongoing support require?

As discussed in previous articles, founders need to think beyond FDA approval. The most successful companies develop an early understanding of who the buyer is, where their technology falls on the priority stack for providers/patients and how the solution integrates into the realities of modern healthcare delivery. Companies that solve workflow challenges often have a significant advantage over those that simply improve technical performance.

Alignment with reimbursement

The final criterion, and often one of the most important factors separating fundable AI MedTech companies from unfundable ones, is whether there is a credible path to reimbursement.

This remains one of the most unresolved challenges in the category. 

We are seeing the establishment and growing usage of AI-focused HCPCS/CPT codes but reimbursement rates remain inconsistent. We still have a long way to go in establishing predictable reimbursement pathways for AI-enabled medical devices.

However, investors still want to see founders actively thinking about what the path to reimbursement may look like based on existing precedents, coverage decisions and coding frameworks, as well as what clinical data will be required to support reimbursement over time.  

Founders are expected to be thinking about the following from day one: 

AI capabilities that aren’t tied to economic value or reimbursement logic are features. AI capabilities that reduce length of stay, lower readmission rates, and improve diagnostic accuracy in ways that qualify for existing or custom CPT codes are businesses.

What this means for founders building AI-enabled companies

The shift from AI as differentiator to AI as baseline expectation has a set of practical implications for founders that are worth stating plainly.

Clinical validation needs to happen earlier. The days of raising on AI potential and deferring the validation work to post-funding are largely behind us. Investors want to see early evidence that the AI capability produces the outcomes the company claims. 

The reimbursement conversation needs to be part of the early logic. For AI-enabled MedTech companies, reimbursement is not a downstream commercial problem. It’s a core business model question that shapes clinical strategy, product design, and capital planning from the beginning. 

Founders who deeply understand potential paths to reimbursement raise from a fundamentally stronger position than those who treat it as something to figure out after FDA clearance.

Workflow integration is a product decision, not a sales problem. The AI companies commanding premium valuations have typically built their products around clinical workflows rather than alongside them. That’s a product architecture choice that gets made early and is expensive to undo. Founders who spend time in clinical settings before they build consistently build products that are easier to sell and easier to scale.

The companies commanding the most serious investor attention in this environment aren’t the ones with the flashiest AI. They’re the ones that can answer the hardest questions about their AI most clearly: What does it do, exactly? Where has that been validated? How does it fit into the clinical workflow? What does reimbursement look like? 

Those are business model questions, and the answers to them are what separate AI-enabled MedTech companies that raise efficiently from those that don’t.

Stifel Venture Banking is a division of Stifel Bank, Member FDIC. For informational purposes only. Stifel Bank does not provide legal, tax, or other advice.