How Generative AI is Transforming the Healthcare Industry?
-
The term 'disruption' is often associated with technological substitution, offering better ways to accomplish specific tasks. But its deeper meaning lies in the transformation of ecosystems—reconnecting and resetting the boundaries of old silos. This distinction is crucial for addressing the impact of generative AI.
Take Napster, for example, the file-sharing system that disrupted the music industry. Before its emergence, music companies debated for years how to engage with digital music. Then, Napster took the decision out of their hands—it broke the deadlock. Initially, industry leaders sounded alarms over rampant intellectual property theft. But ultimately, 'music-as-data' ushered in a new golden age of profits, as new players like Spotify and Apple Music reintegrated individual songs into personalized music streams. The model shifted from album sales to predictable monthly subscriptions. Today, more people listen to more music in more places, and music companies enjoy unprecedented profits. Ecosystem transformation unlocks value. Currently, debates about the impact and significance of generative AI language models like ChatGPT share many similarities with the inflection point triggered by Napster: the astonishing speed of breakthrough technology adoption, the appropriation of others' data (OPD), and predictions of doom and obsolescence. Although OpenAI and ChatGPT, like Napster, may eventually be replaced by later organizations and platforms, the AI revolution they have sparked is irreversible. (While Napster's core value lay in its ability to pirate and distribute others' music, ChatGPT's use of others' data to train its large language models has already sparked numerous lawsuits. It is foreseeable that intellectual property issues will become a major challenge for AI's future. However, our focus here is not on the original training data but on the new proprietary data to which these learning models will be applied.)
The emergence of large language models and other new AI methods will reshape industries—how should leaders prepare? While our discussion centers on AI's impact on the U.S. healthcare industry, our broad perspective applies to every complex ecosystem grappling with this new phase of the digital revolution. As a technology executive and former healthcare system CEO (author Weinstein) and a strategic researcher and consultant (author Adner), we present these insights hoping leaders can envision new approaches to strategy formulation and interaction. Three touchstones will help leaders adjust strategies and prepare for transformation: First, distinguish between AI's role in driving technological substitution and its role in ecosystem transformation. Second, prepare for the challenges of new organizations, which are necessary for ecosystem transformation to realize its value. Third, develop strategies to leverage new asymmetries generated by novel combinations both within and outside the organization.
ChatGPT broke records in technology adoption, reaching 100 million users within two months. Most discussions focus on which jobs it will improve and which it will replace. In other words: technological substitution and how to respond. However, what truly changes the game and presents the greatest opportunities for transformation is its disruption at the ecosystem level. By combining and analyzing data from previously disconnected silos, generative AI creates opportunities to enhance the efficiency and effectiveness of entire healthcare services. Consider three examples: In the United States, administrative expenses account for 15%-30% of healthcare expenditures, with approximately half dedicated to hospital billing and insurance-related costs. Even these estimates are unfair as they overlook the non-monetary indirect costs borne by patients and their families—time spent fighting for insurance coverage and clarifying bills.
Leveraging AI to bridge the gaps between insurers, hospitals, and consumers can automate claims management, prior authorizations, and even payment plans and collections, helping eliminate significant inefficiencies in the system.
Equipment, medications, hospital beds, and staff are often in a state of oversupply (as buffer reserves for emergencies), while sudden shortages (when emergencies exceed expectations) plague the healthcare system. Poor patient flow management leads to unnecessarily prolonged hospital stays and delays in admitting emergency cases. Lack of coordination with continuity-of-care and rehabilitation facilities increases time spent in the most expensive care settings and raises patients' risks of hospital-acquired complications. AI will enable cross-platform coordination among hospitals, systems, partners, and suppliers to build greater resilience and better patient placement, reducing risks, shortening recovery times, and improving outcomes while cutting costs. Avoidable surgeries yielding positive outcomes? Unnecessary tests providing accurate results? These paradoxes highlight the need for quality and performance metrics that more comprehensively consider the patient treatment process. By incorporating the latest advances in medical science and real-world evidence into treatment recommendations and metrics, AI can improve patient outcomes and elevate standards, thereby reducing the burden on both patients and healthcare systems.
When envisioning the future of AI, consider the balance between inside-the-box replacements and cross-silo transformations. How does this balance reflect in your investment priorities: capital expenditures, operational expenditures, and capability development?
Under the current system, successful participants struggle to find a new equilibrium, hindering transformation. High costs on one income statement appear as high revenues on another. These income statements, whether literal or figurative, are determined by organizational boundaries, routines, and records. The aforementioned AI-driven transformations in billing, resource management, and quality all depend on sharing data in novel ways. However, this novelty also introduces a new set of organizational challenges. Historically, the hierarchical decision-making structure guiding reporting within organizations matched parallel information hierarchies—incomplete views across silos might lead to suboptimal decisions but allowed clear decision paths and more efficient execution. This was true within organizations (e.g., nurses couldn't access HR records) and between organizations (e.g., hospitals couldn't access insurers' financial records). However, the benefits promised by transformative AI rely on bridging these silos. This means that, beyond privacy and security concerns, true transformation requires organizations to rethink the informational foundations of power. Once released into data pools, AI eliminates organizational gatekeepers of information. This represents a massive ecosystem shift, moving the focus from ensuring content accuracy ("Is the data correct?") to controlling the breadth of questions ("Who is allowed to ask what?"). The transition from scrutinizing data to scrutinizing questions signifies a fundamental change in the principles of administration and management.
The new visibility enabled by combined data sparks discussions about relevant and appropriate metrics, which in turn influence goals and incentives. Core questions—such as what defines success and who defines it—come into sharp focus. Consider measuring a surgeon's productivity in a world where data can be viewed across silos. Would you count the number of procedures they manage in a month? The revenue they generate? What weight would you assign? In a world of merged data pools and open queries, anyone with access can create their own new metrics, and the system must find a new equilibrium. The visibility across data silos naturally leads to expectations for more comprehensive decision-making that considers a wider range of circumstances. Historically, doctors recommended the "best" treatment based on optimizing medical outcomes. However, with the help of AI, the concept of "best" could change dramatically when considering the broader context of patients' lives—their insurance coverage, work situations, and family circumstances. Incorporating economic and social perspectives into medical advice in a morally and legally defensible manner will become a significant new requirement for healthcare providers and insurers.
As AI improves visibility between silos, organizations need to reassess which parts of their organizational charts and governance structures require adjustment. Proactive planning is essential to ensure that the benefits of broader information access are not overshadowed by unintended drawbacks, such as new sources of conflict. For every organization within the ecosystem, this will necessitate a redefinition of rules and roles. Successful organizations will approach this deliberately rather than reactively.
Ecosystem transformation redraws boundaries. This may require rethinking the historical advantages of scale and scope, but it does not necessarily mean that existing enterprises will fail while new ones rise. Instead, it signifies that the pathways to success will evolve. When scale can be aggregated, size becomes less significant—traditionally, scale and the ability to achieve economies of scale have been interconnected. The potential to leverage shared technology platforms and digitally aggregate the activities of various actors has loosened this coupling. The AI world elevates the promise of aggregation to a new level, allowing information repositories to shift negotiations from a price basis to a delivery effectiveness basis. This creates the potential for decentralization and more equitable distribution of healthcare services, enhancing the viability of smaller medical institutions—a potentially surprising outcome of the current revolution.
New connections drive new synergies—ecosystem transformations align with the expansion of value propositions. This, in turn, resets the logic of scope. By unlocking previously inaccessible internal data, AI can create new synergies within and between organizations, altering the rationale for which activities make more or less sense to undertake together. Over the past 20 years, many healthcare systems have launched their own insurance plans, with mixed results, as the payment and delivery aspects of organizations often operated in silos. AI not only integrates data views across these aspects but also transforms proactive management for both organizations and patients, thereby unlocking value opportunities. Viewing patients holistically—considering health, employment, living conditions, and social needs—provides an opportunity to redefine care and services by reorganizing activities across the entire system. Leveraging insights into these social determinants of health enables targeted prevention and mitigation measures, reducing treatment needs and costs in service delivery while benefiting from cost savings on the payment side, ultimately fulfilling the promise of value-based, profitable care. Those organizations that first unleash their internal synergies will possess a favorable starting point to establish partner alliances for pursuing this new care model – catalyzing a 'minimum viable ecosystem' of key players. This will grant their organizations privileged positions in the quest for ecosystem leadership and reward their efforts.
While obstacles and uncertainties surrounding AI undoubtedly persist – technological, regulatory, ethical, etc. – these can only be resolved through the deliberate efforts of leaders who drive coordinated change, supported by other leaders whose organizations must join this endeavor. For attentive actors, shifting boundaries create opportunities to either lead transformations themselves or support other potential leaders seeking to drive change through mutually beneficial approaches. To better control the ecosystem's destiny, one must simultaneously consider both individual and coalition perspectives. In transitioning to a new ecosystem, what role is my organization truly prepared and willing to play? The decade-long debate over whether to participate in the AI revolution has ended. But how to participate is now the critical question: the choice between being an alliance leader or an advantageous follower is pivotal, with each role presenting its own opportunities and requirements.
Napster catalyzed the transformation of the music ecosystem. Yet, it was only the beginning—not the final act. Without its disruption, it’s hard to imagine a music industry content with its existing structure and profits embracing bold experiments like subscription-based and ad-supported streaming. But that world has arrived, and we’re better for it.
Similarly, ChatGPT is merely the opening act. It’s obvious that those who ignore this new generation of generative AI will face obsolescence. Less obvious is that those who view it purely as a technological disruption risk missing its broader implications. The impact will be shaped by both your ecosystem strategy and the technology itself. Given the urgency and opportunities in the critically important field of healthcare, we focus our discussion here, but this also provides broader insights for businesses, nonprofits, and regulators across economic sectors. Taking an active role in breaking deadlocks, rather than letting others set the terms, is key to navigating the evolving landscape. When properly guided, AI can indeed be a rising tide that lifts many boats. Like the tide, it has already arrived.