Generative AI has crossed a major milestone in American healthcare. A new McKinsey & Company survey, released April 16, reveals that 50% of US healthcare organizations now actively use generative AI. The findings signal a dramatic shift from cautious experimentation to real-world deployment — and the pace is accelerating.
The State of Generative AI in Healthcare
McKinsey surveyed US healthcare leaders during the fourth quarter of 2025. The study covered a broad range of subsectors, including care delivery organizations, payers, and healthcare services and technology firms. Together, the results paint a clear picture: generative AI adoption is no longer a future ambition — it is a present reality.
More than 80% of surveyed leaders confirmed that their organizations have deployed at least one generative AI use case to end users. Additionally, every single respondent reported plans to pursue the technology further. This near-universal commitment reflects growing confidence across the sector.
How Adoption Has Grown Over Time
The growth trend is striking. In late 2023, only 25% of organizations reported implementing generative AI. By 2024, that figure rose to 47%. By the end of 2025, it reached 50%. Consequently, healthcare is now one of the most active sectors for AI adoption globally.
Moreover, the shift is no longer just about piloting tools. Leaders are focusing on scaling and deeper integration into existing clinical and administrative systems. This transition marks a new phase of maturity in healthcare AI.
Top Use Cases Driving Implementation
Administrative Efficiency Leads the Way
The survey identifies administrative efficiency as the area with the greatest potential for generative AI. Tasks such as documentation, billing, scheduling, and prior authorization are prime targets. These functions are time-intensive and rule-based — exactly where AI performs well.
Clinical Productivity Is Also Surging
Meanwhile, clinical productivity has emerged as one of the most widely implemented use cases. More than half of respondents from care delivery organizations reported using generative AI to support clinical workflows. This includes tools that assist with note-taking, clinical summaries, and care plan generation. Therefore, both operational and patient-facing functions are benefiting from AI integration.
Agentic AI: The Next Frontier
Interest is building rapidly around “agentic AI” — systems that go beyond answering questions. These tools can take autonomous action and coordinate complex, multi-step workflows. Currently, 19% of organizations report implementing agentic AI capabilities. However, 51% are already pursuing proofs of concept. This suggests that agentic AI will become a mainstream healthcare tool within the next few years.
Barriers Healthcare Leaders Still Face
Despite strong momentum, adoption is not without friction. Leaders point to several persistent challenges. First, integrating AI into legacy systems remains technically complex. Second, many organizations lack the internal talent to build and manage AI solutions. Third, concerns about AI inaccuracies, bias, data security, and regulatory compliance continue to slow decision-making.
Furthermore, building trust in AI outputs — particularly in clinical settings — requires rigorous validation. Organizations must balance speed of adoption with patient safety obligations. These are not small hurdles, but they are manageable with the right strategy and partnerships.
Financial Returns on AI Investments
Despite the challenges, optimism about financial performance is high. The survey finds that 82% of healthcare leaders expect a positive return on investment from their AI programs. Among those who have already quantified returns, most report gains ranging from less than two times to four times their initial investment. Notably, 45% of respondents have already documented measurable financial results.
These figures make a compelling case for continued investment. Organizations that act now stand to gain competitive advantages in both cost efficiency and care quality.
Strategies Organizations Use to Build AI Capabilities
Healthcare organizations are taking varied approaches to building their AI capabilities. Partnering with third-party vendors remains the most common strategy. However, health services and technology firms are more likely to pursue in-house development. Others are evaluating off-the-shelf solutions as a faster path to deployment.
Each approach carries trade-offs. Vendor partnerships offer speed and proven tools, but they limit customization. In-house development provides control, but demands significant resources. As a result, many organizations are blending strategies — using vendors for quick wins while building internal expertise over time.
Key Takeaways for Healthcare Leaders
Generative AI adoption in US healthcare has reached an inflection point. The data from McKinsey’s survey offers several clear signals for decision-makers:
- Adoption is accelerating — from 25% in 2023 to 50% by end of 2025.
- Administrative and clinical efficiency are the top priority use cases.
- Agentic AI is emerging quickly as the next wave of healthcare automation.
- ROI is real — 82% of leaders expect positive returns.
- Barriers remain, but they are addressable through planning and partnerships.
For healthcare leaders still on the sidelines, the window for early-mover advantage is narrowing. The organizations building AI fluency today will likely define care delivery standards tomorrow.
