Vendors are slapping an AI label on tools that have not changed and charging health systems far more for them, and CIOs say they are not paying for the label alone.
Health system CIOs have grown accustomed to vendors adding artificial intelligence features to existing software. What has become a sharper point of frustration is what often comes with it, a renewal quote several times higher than the year before, attached to a tool whose underlying workflow has not meaningfully changed.
Muhammad Siddiqui, CIO of Richmond, Ind.-based Reid Health, said the pattern is familiar, and common enough that it prompted collective action. Earlier this year, he joined a group of peer health system CIOs in signing a letter to the CEO of a major software vendor, raising the issue directly.
“AI pricing is running ahead of AI value,” Mr. Siddiqui told Becker’s.
The pattern he described, tools the health system has run for years returning at renewal with an AI label and a price several times higher while the core workflow stays the same, has shown up most often in back-office and productivity software, according to Mr. Siddiqui.
Tom Bartiromo, senior vice president and CIO of West Reading, Pa.-based Tower Health, said he has not seen price jumps of that scale, but the underlying trend is familiar. Established software capabilities are increasingly repackaged as AI, he said, often paired with new consumption-based fees that can push the total cost several-fold higher once real transaction or call volumes are applied.
“The concern is whether the product delivers a genuinely different outcome or is simply existing automation with an AI premium,” Mr. Bartiromo told Becker’s.
Not every CIO has seen the same magnitude of increase. Darrell Bodnar, CIO of Berlin, N.H.-based North Country Healthcare, said he has encountered modest price increases and situations where an existing product was repositioned around a new AI feature, but has not seen a tool jump five- or tenfold simply because it was rebranded as AI. Still, he said, any significant increase requires clearly demonstrated return on investment before he will approve it.
“An AI label alone would not justify the additional cost,” Mr. Bodnar told Becker’s. “Regardless of the presence of AI or not, I always perform due diligence.”
The caution echoes a broader shift Becker’s has reported across health system IT departments, where leaders including those at Houston Methodist have begun restricting AI tool access and spending until a clear return is proven, rather than approving new licenses by default.
That default has shifted at Reid Health, Tower Health and North Country Healthcare alike. All three CIOs described a similar test before agreeing to pay more, a measurable result in their own environment, verified against a baseline set before any AI tool goes live.
“My bar is a measurable outcome in my environment, not in a vendor slide,” Mr. Siddiqui said.
His standard requires time given back to clinicians or dollars saved, measured against a baseline set before a pilot begins, with an exit clause if the results do not hold. Reid Health’s ambient documentation platform met that bar, he said, after the system measured time returned per encounter and note quality against its pre-AI baseline.
“A relabeled feature at a multiple of last year’s price,” he added, “has never met it.”
Mr. Bartiromo requires results validated through a proof-of-value engagement under real operating conditions rather than vendor benchmarks alone, citing improved accuracy, reduced labor or cost, a better patient or employee experience, and acceptable safety and reliability.
“A demonstration is not evidence,” Mr. Bartiromo said. Some AI solutions can meet that standard for targeted use cases, he said, but broad enterprise claims frequently do not survive real-world complexity.
When a vendor cannot clear that bar, both CIOs said they would rather keep what they already have.
“If the AI does not change an outcome for our patients or our staff, the older version at the original price is the better buy,” Mr. Siddiqui said. “An AI upgrade we did not ask for is a price increase, and we negotiate it like one.”
Mr. Bartiromo said he would make the same call.
“If the AI version does not produce a material, measurable improvement, I would keep the proven non-AI product at the original price,” he said. “‘AI’ is not itself a business outcome. An upgrade we did not request should not become a forced price increase or ‘AI Tax,’ particularly when the customer is also being asked to absorb implementation risk, model variability, and ongoing consumption charges.”
Mr. Bartiromo said the responsibility ultimately sits with vendors as much as with the health systems paying them.
“Vendors seeking a premium need to share the risk by tying a meaningful portion of their compensation to the outcomes they promise,” he said.
That, according to Mr. Siddiqui, is the argument he and other CIOs have already begun making directly to vendor leadership, and one he expects more health systems to make as AI pricing continues to outpace proof of what it delivers.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.