AI strategy is pulling health system C-suites in opposite directions as technology evolves. They’re focused on moving quickly where value is clear, but slowing down to achieve governance, fiscal discipline and ethical oversight. Sometimes they’re aggressive to invest and roll out AI, while other times they’re more cautious and pulling back.
The dichotomy might seem chaotic, but there is a method to the madness.
Patrick O’Shaughnessy, DO, president and CEO of the Rockville Centre, N.Y.-based Catholic Health, said the pace of AI development has created a dynamic healthcare leaders have not faced before. Unlike past technology waves, where adoption lagged innovation, AI is advancing faster than many organizations’ ability to manage it responsibly.
“It’s kind of interesting,” Dr. O’Shaughnessy said during a recent “Becker’s Healthcare Podcast” interview. “We’re actually now saying, ‘Hey, wait a minute. We need to slow this down just a touch. Make sure we have good governance, good control over it.’”
That shift has shaped Catholic Health’s overall AI philosophy. Rather than viewing AI as a replacement for human labor or judgment, the system is positioning it as a tool to strengthen decision-making and performance across the organization.
“Our strategy is more augmented AI – how do we help our people perform better – not necessarily autonomous AI where we’re replacing people,” said Dr. O’Shaughnessy.
Governance as a guardrail against AI sprawl
To operationalize that philosophy, Catholic Health has put formal governance structures in place to manage both existing technology and emerging AI opportunities. The goal is to ensure AI adoption remains aligned with strategy, mission and financial realities.
“We’ve built an IT executive board that actually will govern not just a day-to-day ‘lights on’ operations, but then also innovation and an innovation subgroup that actually meets where AI lives for us and really will help us guide the right decisions for the health system,” said Dr. O’Shaughnessy.
Within that framework, leaders evaluate AI initiatives through a practical lens, focusing on sequencing, purpose and value rather than speed alone. The governance body is designed to bring discipline to what could otherwise become an unmanageable volume of requests and pilots.
The governing body decides which projects to optimize, when to buy best of breed, what and when new technology roll-outs will be based on the most beneficial projects for the system. They identify the right problems to solve with technology and manage resources appropriately. Without that structure, the demand for new AI tools can quickly overwhelm executive teams and strain operating budgets.
“Without that governance structure, I think it’s a little bit of the wild west,” he said. “Certainly people in my position, my CFO, my COO, would get inundated with requests for best of breed spin off technologies, and it quickly becomes unsustainable.”
Avoiding a ‘Frankenstein’ tech stack
Beyond governance, Catholic Health’s approach is also informed by financial discipline and long-term sustainability. As vendors race to bring AI-enabled products to market, Dr. O’Shaughnessy said health systems must resist the temptation to layer on disconnected tools that increase complexity and cost.
“So many of these companies and we work with our payers and we work with independent spin offs that have best of breed type bolt on technologies,” he said. “But if you’re not careful, you wind up building a Frankenstein that does not necessarily achieve your mission objectives, and it costs you a lot of money to just install it and then to maintain it with all the operating agreements.”
Instead of defaulting to new purchases, Catholic Health is prioritizing optimization of existing technology investments, particularly its EHR, before adding external solutions.
“We’re taking a combination approach and it’s mixed,” Dr. O’Shaughnessy said. “It’s aggressive, but also at the same time cautious and trying to be fiscally responsible as to using what we’ve already implemented and turning on those switches and activating it.”
Purpose first, technology second
AI governance is a financial or operational issue and a question of ethics, mission and outcomes. As AI moves deeper into clinical and administrative workflows, leaders must be clear about why technologies are being deployed in the first place.
“Make sure you have the right people around the table,” Dr. O’Shaughnessy said. “People that are not afraid to say, ‘Hey, wait a minute. We need to pump the brakes here. This is something that doesn’t look right, feel right’.”
The slowdown will help refocus the project and guard against misaligned investments. If the AI platforms and applications don’t meaningfully improve operations or patient care, then it’s not worth the system’s time. Dr. O’Shaughnessy and his team reflect on several questions ahead of any final AI decisions.
“Are we really bending the disease care curve?” Dr. O’Shaughnessy said. “Are we getting to patients sooner in their disease processes? Are we delivering the best in quality, evidence based care quicker for them, getting them down their diagnostic path sooner? That is our goal.”
