Leveraging AI for Strategic Decision-Making in Healthcare Operations

Abstract 

Strategic decision-making is critical for healthcare leaders due to its impact on both organizational success and patient outcomes. However, strategy is complex. This article offers strategies addressing that complexity by leveraging AI to enhance strategic decision speed and quality across healthcare organizations. While AI adoption in healthcare is growing, its use in executive decisions needs additional practical guidance. We highlight five main ways AI can benefit strategy development in health organizations: operational efficiency, identifying opportunities, market forecasting, risk management, and optimizing strategies. Generative AI and large language models can process large amounts of information and assist with tasks like scenario planning, market research, and due diligence. We also examine how aggregating diverse AI-generated insights can help healthcare leaders substantially reduce the bias and inconsistency seen in single-model outputs. This combining of multiple assessments yields evaluations similar to expert judgment, making AI a valuable support tool for healthcare executives.

Introduction

Strategic management is the process of implementing an organization's strategies to accomplish objectives established for its principal stakeholders.[1] Strategic decision-making is perhaps the most challenging and consequential of all responsibilities held by senior leadership. Moreover, the healthcare sector is among the most complex and impactful sectors, with life and death often literally involved. Thus, when combined, strategic management for the healthcare industry is particularly complex, consequential, and challenging. 

 

AI is increasingly being used in healthcare to improve patient outcomes (e.g., access, diagnosis, treatment), but the utilization for overarching organizational strategy is lagging. Per a recent article in Sloan Management Review, there is a “A Tidal Wave of Adoption, a Trickle of Strategy”, with many organizations adopting agentic AI before they have a strategy in place.[2] Thus, the foundational strategic thinking around AI needs to catch up with both the potential of AI and its increasingly widespread day-to-day utilization. In this article, we provide conceptual guidance to healthcare leaders on how AI can be used to improve strategic decisions in healthcare. 

 

Strategic Management in the Traditional Approach

Strategic planning sets an organization's vision, goals, and approach to reaching business objectives. It typically uses frameworks (e.g., Balanced Scorecard, BCG Matrix, Porter’s Five Forces, SWOT) and data to gain insights, create options, make decisions, and execute actions. These decisions are uncertain and often cannot be reversed, so accurately estimating the value of alternatives is crucial. Moreover, strategic decision making frequently takes place when data are scarce and situations are unique.[3]

Traditional scenario planning tools and methods often lack the speed and breadth required to address the complexity of healthcare business environments. Given the current scale of available data, conventional forecasting and market analysis methods are increasingly inadequate. These methods are limited by (i) dependence on past data, (ii) lengthy analysis, (iii) relatively few variables considered, (iv) low and slow adaptability to real-time changes, and (v) susceptibility to human bias. Strategic planning has evolved beyond defining general objectives, and it now relies on data-driven decision-making to achieve measurable outcomes. Many companies are leveraging artificial intelligence and machine learning to enhance their strategic planning processes. This provides a significant advantage by efficiently processing large datasets and delivering actionable insights.

 

Advantages of AI in Decision-Making

Artificial intelligence can be expected to play an increasingly integral role across all aspects of healthcare decision-making processes. It can help uncover challenges and opportunities that may otherwise go unnoticed. This technology is most valuable in the design phase, enabling healthcare organizations to evaluate their position within an industry. Advances in AI provide new ways to assess strategic choices like business model selection, acquisitions, and organizational redesign.[4] Per this potential, McKinsey & Company recommends three immediate actions for senior leaders: (i) learn the fundamentals of AI, (ii) begin building now, (iii) create your own insights ecosystem.[5] This learn/build/create now approach is fully applicable to healthcare organizations. 

AI will not replace human reasoning in strategic domains, but it can provide objective answers that enhance decision-making. As AI takes on roles like researcher and thought partner, it may redefine how strategists work and support strategic choices. By streamlining strategy development and leaving room for creative ideas, AI offers leaders a competitive advantage. There are five key ways AI can help CEOs make better decisions: enhancing operational efficiency, identifying opportunities, market forecasting, risk management, and strategy optimization.[6] AI-powered analytics enable more accurate market forecasts, better opportunity detection, and improved strategy optimization. Instead of relying on intuition or limited data, healthcare leaders can use AI-infused strategy to make informed, agile decisions to keep their organizations competitive. 

Large language models use deep learning with neural networks. "Large" refers to their huge parameter count, sometimes exceeding one trillion. Current LLMs are already capable of handling large-scale synthesis tasks, such as market research and can perform some data aggregation and reasoning. One study found that OpenAI’s basic model (utilized without fine-tuning) performs reasonably well compared to a leading candidate in consulting case interviews and is capable of generating a substantive initial draft for buyside due diligence analyses.[7] Accordingly, LLMs can already serve as effective assistants in strategy-related tasks, a development not observed just a few months ago, which captures how rapidly this technology is improving. 

 

AI can Transform Business Planning

There are five specific ways AI is transforming scenario planning.[8] First, by integrating both structured and unstructured information, it promotes comprehensive data integration and enables consolidation of disparate data sources (e.g., financial records, operational metrics, patient sentiment). Second, it quickly creates detailed scenarios within minutes, eliminating the need for extended spreadsheet modeling and iterative revisions. Third, it promotes adaptive modeling by dynamically incorporating new data from economic trends, regulatory changes, and competitor activities, automatically updating models as conditions evolve. Fourth, it promotes evaluation of atypical scenarios by simulating edge cases and low-probability events that traditional methods may fail to consider. Fifth, it promotes enhanced communication as AI-powered visualization transforms complex data models into clear narratives. This allows decision-makers to gain insights into financial, operational, and workforce impacts through tailored presentations such as dashboards, charts, or story-driven formats.

AI acts as a (i) researcher, (ii) interpreter, (iii) thought partner, (iv) simulator, and (v) communicator on healthcare strategy.[9] As a researcher, AI collects and enhances data from a wide range of sources. As an interpreter, AI interprets data analytics results to understand how findings can help the healthcare organization achieve objectives. As a thought-partner, AI acts as a brainstorming team member, accelerating idea generation and helping business leaders overcome biases. As a simulator, AI systematically assesses different market scenarios, including economic shifts, competitor activity, and stakeholder reactions. As a communicator, AI outlines the strategic effects on the organization and its stakeholders, presenting this in a variety of ways (e.g., dashboards, infographics, and PowerPoints).

The data sources used by these GenAI systems combine information from financial records, news articles, official regulatory documents, and customer opinions to create comprehensive scenario inputs. This GenAI can be applied in multiple specific ways.[10] First, with forecasting and scenario modeling, generative AI uses historical and external data to simulate business scenarios, helping CEOs test strategies and forecast outcomes under different conditions. Second, with dynamic strategy generation, generative AI can create strategic plans tailored to businesses and adapt them in real time based on new data and feedback. Third, with competitor and market simulations, healthcare executives can use AI to model competitor behavior, forecast market changes, and refine their strategies. Fourth, with specifically tailored business solutions, GenAI helps leaders design custom business models, marketing plans, and operations that directly target specific company challenges and opportunities. Fifth, with automated decision-making, any routine strategic decisions can be automated with pre-set parameters and real-time data. 

 

Best Practices for AI Governance

There are some best practices for AI governance. These include setting standards (define AI policies for planning), monitoring data (keep inputs accurate and protect privacy), reviewing results (have experts validate AI outputs), and integrating effectively (align AI with current planning processes). When AI systems feature transparent decision-making, they are easier to audit and explain to stakeholders such as customers, regulators, and internal teams. 

As part of effective governance and utilization, there should be descriptive tools providing advanced machine learning for tasks such as customer data analysis, customer segmentation, demand forecasting and operational optimization to help make informed, efficient decisions. Second, there should be predictive/prescriptive tools for building models, analyzing data, and generating insights. It should be suited for businesses aiming to use AI in decision-making, including forecasting, risk analysis, and market trends. Third, there should be a data visualization platform with AI features like “Explain Data” and “Ask Data” to pose questions in natural language and get instant answers.

Weak AI governance can expose healthcare organizations to serious risks, including breaches of patient privacy, regulatory violations, and unchecked algorithmic bias. These failures may result in investigations, costly fines, or even operational disruptions. Beyond regulatory and legal risks, governance lapses can quickly erode stakeholder trust and damage reputation. Patients, partners, and investors may lose confidence, leading to financial losses as business relationships suffer. To prevent these outcomes, healthcare executives must ensure AI systems are transparent, auditable, and regularly monitored for risks like data quality and explainability. Proactive governance protects organizations and supports a culture of accountability and innovation.

 

Healthcare Leaders Still Needed

AI is a powerful aid, but because strategy involves such complexity, human input will remain integral for the foreseeable future.[11] Strategy is challenging due to the need for multi-level reasoning, contextual awareness, and understanding of human behavior.[12] These models cannot yet perform complex, hypothesis-driven, multi-step reasoning. Healthcare institutions face many intricate analyses (e.g., buyside due diligence on a major clinic acquisition, determining whether to close a hospital, or deciding whether to significantly expand a care unit). Careful review by humans will still be required to validate calculations and ensure that any incorrect assumptions made by the model are identified and amended.[13]

When used to help humans, AI poses risks such as data non-representation/skewness, lack of clear explanations, and producing false but convincing information (i.e., hallucinations). In particular, generative AI models exhibit inherent biases associated with the datasets and natural language tasks utilized during their pre-training processes. Thus, it is important to review the diversity and representativeness of the training datasets used by these systems. Systems trained on biased or incomplete data can generate discriminatory or inaccurate results, leading to poor decisions. Ensuring high data quality through proactive measures prevents costly mistakes and improves system reliability. 

In terms of on-going human engagement and supervision of AI, there are key points for organizational leaders to consider.[14] First, proprietary data access becomes more critical. Second, increased data makes separating valuable information from noise essential. Third, with easier insight generation, executive-level synthesis gains importance. Fourth, strong strategy development processes matter more than the quality of insights. Fifth, strategy teams must invest in technology to build and access proprietary data ecosystems. And in the final analysis, it will be the responsibility of healthcare leaders to make the ultimate difficult strategic decisions. As McKinsey notes, “AI won’t change the need for leaders to demonstrate strategic courage by committing to big moves.”[15]

 

Power of Averaging Multiple Perspectives

Strategic foresight, or predicting the results of strategic decisions, is central to key strategy theories.[16] Researchers have focused on how both individual and combined predictions influence the evaluation of a strategic decision.[17] Prior research has shown that aggregating many imperfect predictions can improve the overall prediction by offsetting errors.[18][19] The benefit of aggregation is also often called wisdom of the crowds.

Previous research on human evaluators has examined the effects of differences in their expertise, cognitive approaches, and demographic characteristics.[20] Aggregating multiple predictions improves accuracy when their errors partially cancel out. In regression tasks, positive and negative errors balance each other, while in classification tasks, majority voting ensures most correct predictions prevail. Aggregation works best with a larger and more diverse set of predictions. It can occur through many techniques, including by averaging, majority voting, or taking a modal hybrid. A widely adopted method involves selecting evaluators with varied backgrounds. 

Researchers compared rankings from large language models versus human experts in analyses of 60 business models.[21] They find that generative AI can produce inconsistent and biased evaluations, but its aggregated rankings are similar to those of humans. This study shows that generative AI offers useful predictions for strategic decision making. Single evaluations from generative AI can be inconsistent or biased. However, combining multiple assessments from different LLMs, prompts, or roles produces results similar to human experts. This method efficiently offers healthcare executives strategic insights across domains and can supplement human judgment. 

When applying the wisdom of crowds, we need both scale and diversity in predications.[22] First, diversity means predictions vary from one another. By combining different predictions, errors can be balanced out, so optimistic forecasts offset pessimistic ones for continuous outcomes. Without diversity, group predictions offer little advantage over individual ones. Second, scale is the number of predictions used in aggregation. Using many predictions increases the chance of offsetting errors. If too few predictions are chosen, they may all be overly optimistic or pessimistic, reducing aggregation effectiveness.

 

 

 

Conclusion

The rapid integration of generative AI offers healthcare leaders a significant opportunity to enhance strategic planning and decision-making amid increasing complexity and uncertainty. It facilitates forecasting, scenario modeling, competitor simulation, and strategy formulation, enabling healthcare executives to evaluate assumptions and adjust strategies in real time. AI can synthesize a wide array of data sources (such as financial metrics, regulatory guidance, market intelligence, and stakeholder input). These can then convert to actionable insights that inform flexible, evidence-based leadership. 

To fully realize these benefits, disciplined implementation by health systems and hospitals is essential. An effective AI-supported strategy requires both sufficient scale and diversity in predictions. Diversity ensures that AI-generated perspectives are meaningfully varied, which allows errors to offset one another and mitigates systematic bias. Scale amplifies this effect by increasing the reliability of aggregated insights. Absent adequate diversity, aggregation delivers limited benefit; without sufficient scale, predictive models may share similar blind spots. It is therefore crucial for healthcare executives to move beyond single-model outputs and intentionally cultivate a range of independent AI perspectives across different models, prompts, and roles.

When applied in this manner, generative AI serves to augment, rather than replace, executive judgment. Aggregated AI insights can improve strategic foresight, support comprehensive risk management, and elevate the quality of complex decision-making, all while maintaining human accountability and ethical oversight. As healthcare organizations continue to embed AI into their operations, those embracing a thoughtful, aggregation-driven approach will be better equipped to navigate uncertainty, sustain competitive advantage, and enhance both organizational performance and patient outcomes. This will benefit patients, their families, clinicians, other stakeholders, and the organization itself.