Every major healthcare company, along with the smaller ones, are pushing forward on artificial intelligence strategies, and this was clearly evident at recent industry conferences.
If generative AI was the key theme last year, then AI agents were the star of the show at the latest big conference in Las Vegas. During interviews at HIMSS 2025, healthcare executives shared their thoughts on the evolution of AI with a big focus on scaling the technology, integrating it with existing systems and gaining adoption.
Agentic AI also was a buzzword at this year's HIMSS. Designed to be autonomous, agentic AI makes independent decisions and takes action based on the goal it's built around, which goes a step further from AI agents that are designed to handle specific tasks.
"My prediction on the buzzwords of the year are really open-source models and agentic AI," said Don Woodlock, head of global healthcare solutions at InterSystems, during an interview at HIMSS 2025. "Open-source models I think will be just interesting from a PHI (personal health information) point of view. You know, running the model inside your firewalls versus external and I think that will open up AI to some organizations that are a little nervous about it."
"I think agentic AI will be interesting for the next couple years, which is basically allowing the software to do higher level things like, 'This patient's coming in for surgery, can you just take care of the pre-visit appointment, sending the patient some education materials, scheduling the resources,' so having a bundle of activities that can be done for you, which I think will take AI to the next useful level."
EHR companies investing in AI
At HIMSS, InterSystems debuted a new electronic health record and healthcare information system designed with AI and interoperability "at its core," executives said.
At the HLTH 2024 conference back in October, along with ViVE in February and HIMSS this month, the showroom floor was buzzing with startups offering AI solutions.
"I am a little concerned about all the smaller companies because they're not all going to survive," Woodlock mused. "I think that we've proven that integrating into larger platforms like Epic or Cerner or our EHR and HealthShare product line, integrating into the platform is the right place for AI rather than having something off to the side."
He added, "Even the ambient players, I worry about their long-term viability because you really should build that right into the core workflow systems, from my point of view, then you can get the real good integration that you want. I do think some customers will be left with problems that they need to switch vendors a few years down the road."
InterSystems is a nearly 50-year-old company that offers a data management platform for healthcare along with interoperability and data analytics solutions. Its EHR solutions, including its TrakCare EMR, are sold outside the U.S.
The industry also is waiting to see how the big EHR companies will integrate and roll out AI solutions, executives said at HIMSS. Epic is putting major investments into rapidly building out its AI capabilities including conversational AI features.
"We've woven AI into the foundational capabilities of Epic, and we've been working towards an agentic platform for the past year or so," Seth Howard, vice president of R&D at Epic, told Fierce Healthcare during an interview at HIMSS.
InterSystems is taking a similar approach by building AI capabilities within its EHRs rather than adding other companies' solutions.
"We're building ambient ourselves right within our application because we feel like we can do the best integration that way. I would think that Epic, Cerner and others are going to do the same eventually. The integrated approach is the best versus gluing together multiple systems and with so many smaller companies, I think CIOs aren't going to be excited about gluing together multiple systems," Woodlock said.
How AI strategies will evolve in 2025
Major players in the health tech industry are quickly building AI agents specifically for healthcare. Microsoft rolled out a new AI assistant for healthcare professionals that it bills as an all-in-one technology that combines voice dictation, ambient listening and gen AI. The new tool, Dragon Copilot, combines the natural language speech recognition capabilities of Dragon Medical One with the ambient listening of DAX Copilot, which it launched about a year ago.
Back in February, Innovaccer unveiled a suite of pretrained AI agents that are voice-activated and can communicate with patients for appointment scheduling, protocol intake, managing referrals and answering routine patient inquiries. The suite of eight AI agents also tackles tasks like prior authorization, care gap closure, hierarchical condition category coding and transitional care management.
During a walk around the massive HIMSS exhibition floor, Guillaume de Zwirek, co-founder and CEO of digital health company Artera, pointed out the ubiquity of AI agents, judging by companies' branding and signage. AI agents seem to be everywhere in healthcare now.
"With the proliferation and commoditization of AI across the market, differentiation really comes down to three key things—integration, distribution and documented workflows," he said. "As a trusted partner to over 900 providers, we’ve integrated AI into our platform for years to help them make smarter decisions and drive organizational growth. Artera will continue to build on these advancements to create new AI tooling which drives industry progress and enhances the patient experience.”
Major health systems and provider groups have spent the past year or two testing out AI solutions, including AI scribes to help clinicians with administrative tasks.
"I think we're kind of reaching the phase where there's real value, there's real desire to adopt, in fact, it's kind of not just a push, but a pull. Even the clinicians are kind of driving the flywheel," said Ian Shakil, chief strategy officer at artificial intelligence and software company Commure, during an interview at HIMSS. "Now I think the conversation is evolving to how do we scale and how do we get and maintain utilization, and get and maintain all of these good scores on ROI and delight at these increasingly higher and higher levels of scale."
Shakil added, "I also think AI at a high level, maybe one or two years ago, was really more focused on sort of passive tasks, like the note is being done right over here, where some admin thing is unfolding over here. In 2025, I think we're now looking at more interventional or active uses of AI that are affecting decisions at the point of care, providing tools and reminders to help you think of things differently, take a different action, and that's really exciting."
Commure continues to build out its technology stack to service healthcare organizations with AI-powered solutions for patient engagement, clinical documentation, revenue cycle management and real-time locating systems. The company acquired Augmedix, a healthcare AI company, last July. Shakil was the founder, director and chief strategy officer at Augmedix before moving over to join the executive team at Commure.
The company is now working with HCA Healthcare to codevelop an ambient AI platform to support critical workflows in emergency departments, hospitalist operations and ambulatory care.
Aashima Gupta, global director of healthcare strategy and solutions at Google Cloud, agrees that healthcare organizations have shifted their focus from piloting or testing AI to developing strategies for scaling AI solutions across the enterprise.
"Scaling AI is hard, and as different teams are building different AI, oftentimes that has been left to IT to manage everything," Gupta said during a sit-down at HIMSS 2025. "Now the conversation is, 'How do I lay down the foundation that empowers different teams within my organization to build AI or gen AI capabilities, but do that with enterprise-grade security, privacy, guardrails and prompt engineering?' They are establishing that kind of platform, so moving from one use case to a series of use cases, solving for a business problem, but laying down the foundation so different teams can be more productive."
"This moment for gen AI looks different because healthcare used to be a technologically laggard industry, but I'm seeing organizations embrace generative AI for these type of use cases [documentation and administrative tasks]," Gupta noted.
Provider, payer and digital health organizations also are evaluating the highest ROI area to apply AI solutions, she noted. "Organizations are looking at, 'Where am I getting the huge lift?' Those conversations are progressing."
She added, "As we talk about generative AI and how it can help alleviate some of the manual tasks, the reading of the documentation, writing the documentation, and as it moves from buzzword to really a business essential, we're seeing two trends playing out—AI-powered search and agents."
At the conference, Google's cloud division announced new gen AI capabilities in its Vertex AI Search for healthcare, including a new feature called Visual Q&A that searches tables and diagrams. Medical information isn't just captured in text, but also images, sounds, videos and charts.
"Agentic AI is a significant shift where gen AI was used for search summarization and where we are seeing agentic AI going is that agents can complete or follow up a multiple-step process towards a goal, they act on your behalf," she said. "Today we have agents doing one thing really well. In the future, we're going to have a multi-agent kind of enterprise that provides that orchestration."
AI to augment a human workforce in short supply
Salesforce also joined the race to roll out healthcare AI agents. Prior to HIMSS, the company released Agentforce for Health, a new library of prebuilt agent skills and actions designed to tackle time-consuming administrative tasks in healthcare. The AI tools can check eligibility, schedule appointments, verify insurance benefits and prior authorization and analyze clinical trial sites, the company said.
Amit Khanna, senior vice president and general manager for Salesforce Health, contends that AI agents have the potential to help the healthcare industry face a growing labor shortage by augmenting the work of human professionals and clinicians.
"Healthcare is an industry that doesn't have enough humans to complete the work that it has today," he said.
Moving forward, the healthcare industry should focus on "trusted, in-the-flow-of-work AI that is easy to implement," he said.
"We don't want to be in scenario where we have to hire 200 Ph.Ds. and a large language model because that will go against the $3.7 trillion we already spend in healthcare, we'll just add another $1 trillion to it. We have to get to simplicity of implementing," Khanna said.
Compared to a year ago, when many healthcare executives were trepidatious about implementing AI and the risks involved, healthcare leaders are developing AI strategies and having practical conversations with chief financial officers and compliance officers about AI solutions, Khanna noted.
"One customers told me that he goes to his compliance officer and says, 'This AI is an intern. Think of it that we have hired an intern from college. They need observation. They need somebody to take care of them and tell them what to do. They'll not be on their own, but I'll be managing that intern. Yes, some mistakes will be done, but it's managed by a human.' That's how they are winning the argument with compliance," Khanna said. "Our customers are looking at, how can I remove administrative and operational work to give time back to patient care, whether it's care coordinators, whether it's even call center people. There are certain tasks that a human doesn't need to be doing. I see the conversation has evolved to do it now rather than waiting," he said.