According to Gartner, 55% of companies have deployed AI-first strategies. With AI dominating every business conversation (and beyond), this comes as no surprise. Yet, as with every major technological shift, strategy sometimes lags behind execution, with many simply rushing to stay ahead of the curve. The result? AI pilots that stall, unclear ROI, and frustration in boardrooms everywhere.
Shifting away from AI, however, means falling behind, creating a cycle of reinvesting in initiatives that fail to deliver. It’s increasingly clear that without a clear vision, well-defined use cases, and a comprehensive, ROI-focused strategy, AI projects are unlikely to succeed.
Many of these challenges are not unique to a single industry. Sharing successes and lessons learned across sectors helps leaders see what works, what doesn’t, and how to scale AI initiatives effectively. To leverage collective expertise for broader impact, HTEC has organized a series of AI-first executive panels across the US and Europe, including New Jersey, London, and, more recently, Munich and San Diego.
Let’s explore the topics that dominated the discussions in California, opened by Lawrence Whittle, HTEC’s Chief Strategy Officer, moderated by Alfred E Olivares, HTEC’s Managing Partner for Healthcare and Life Sciences, and featuring esteemed panelists: Shola Oyewole (United Therapeutics Corporation), Rachael Son (Boston Scientific), Julien Willard, MD MPH (Arcstone Private Intelligence), and Tim Sears (HTEC’s Chief AI Officer).
What makes a business AI-first?
Introducing the panel, Lawrence emphasized the importance of escaping the trap of equating AI with cost-cutting. Instead, it’s about increasing throughput – amplifying efficiency and accelerating value creation across the enterprise.
Before defining an AI strategy, however, companies must decide what AI-first means for them. While the panel showed there’s overlap, ultimately, how you define it depends on where you sit.
Shola emphasized the role of data: “As a company growing and 3D printing human organs, we need massive amounts of data. For me, becoming AI-first means relying on ample high-quality data to inform the design of everything you are doing – be it organs, devices, or clinical trials.”
Transforming the patient care space through data-driven insights and value-based decision making took a central spot in Rachael’s definition of an AI-first approach. Such transformation must encompass the entire patient journey, from accurate diagnosis to prescribing personalized, highly targeted treatment to ongoing monitoring for chronic patients.
Although it was somewhat difficult to come to a unified definition of what constitutes an AI-first approach, panelists were unanimous in identifying the major obstacle to achieving an effective AI strategy: frustration from failed pilots.
Too many pilots, too little progress
Parallel with a growing emphasis on combining intelligent AI agents with rich data ecosystems to automate and augment workflows, there’s also rising frustration that heavy investments in AI transformation haven’t paid off. Unclear use cases, over-regulation of the HLS industry, and talent gaps compound the issue of scaling pilots successfully. Businesses must start from a clearly defined problem space and make sure that the problem at hand can be addressed with AI.
“There’s a lack of focus when it comes to scaling AI. Instead of identifying the top 2-3 use cases that can yield a very tangible impact, companies tend to run dozens of pilots in parallel. This approach not only eats up resources without a tangible ROI, but also causes a significant drop in workforce morale, as the outcome is not proportionate to the effort invested,” observed Julien.
This is not limited to the HLS industry – and Tim believes that cross-industry collaboration, where companies share both their wins and setbacks, can help validate approaches, refine strategies, and accelerate scalable AI adoption. In addition, stakeholders must be guided by ROI, rather than the fear of missing out and “doing AI for AI’s sake”.
Two approaches to making an impact with AI
As Rachael pointed out, there are two primary ways companies are approaching AI adoption: evolution and revolution.
“Both ways work, but one must test and see what works in their particular context. For me, it is usually about evolution. Getting to the ‘low hanging fruits’ first to showcase the immediate wins and secure the buy-in across stakeholders who might be at different levels of technical orientation or risk aversion.”
In addition, it’s crucial to ensure that the changes being introduced do not disrupt current habits and workflows; otherwise, they may never be adopted. Shola advises running your AI projects in parallel. That way, you can keep showing the relevance of what you’re doing in comparison to the current workflows and processes, so teams can clearly see the added value AI brings without feeling their existing ways of working are being replaced.
Beyond technology: The change management imperative
An underlying theme of each of HTEC’s AI panels has been the need for transformation on multiple levels: process, workforce, and product/technology.
“Stakeholders are often eager to integrate AI productivity tools and launch AI-driven products, but they sometimes overlook the equally critical transformations in processes and workforce – areas that are essential to ensure AI initiatives deliver real, sustainable impact,” Tim said.
There was a clear consensus among the panelists that people must be actively engaged in the change process early on. As everyone pointed out, resistance to change is a universal human trait – and AI is no exception. People need to be involved in the change, rather than being passive observers. Often, they cling to slow, manual processes simply because they are familiar. By gradually introducing them to new workflows and making the technology familiar, organizations can earn their trust and secure genuine buy-in.
From diagnostics to automating medical writing to personalized treatment, we’ve just begun to scratch the surface of what AI can achieve in the healthcare and medical device space. Yet, laying a solid foundation through clear strategies, transformed processes, and workforce engagement will separate the leaders from those who fall behind. Proving ROI early on turns pilots into momentum, letting success ripple across the organization.
As progress accelerates, it’s clear that no single company or sector can drive this change alone. Collaboration across industries, disciplines, and regions will be key to unlocking AI’s potential in healthcare and beyond (one of HTEC’s major motives to also join the World Economic Forum’s (WEF) Centre for AI Excellence).
We look forward to continuing these cross-industry discussions and thank all the panelists for sharing their insights and experiences.
