The 2026 Revenue Cycle Year in Review: 7 Shifts and What to Fund in Your 2027 Budget
With about ten weeks left in 2026, many health systems are finishing their 2027 budgets. Before yours is final, it's worth looking at what this year changed in the revenue cycle.
I've written about revenue cycle on this blog nearly every week since February. Looking back across those posts, seven shifts stand out. For each one, I've summarized what happened, what it means for 2027, and the budget line I'd fund, cut or protect because of it. There's a list of questions for your finance committee at the end.
One theme runs through all seven. In August, I graded my own RCM 2030 predictions, and every time I was wrong, I was wrong because things moved faster than I expected. If your 2027 plan assumes 2026's pace, it is probably too slow.
1. The AI billing war became public
What happened. In March, I wrote about the AI arms race in healthcare finance, when the Blue Cross Blue Shield Association first argued that AI-assisted coding was driving up costs. In April, I wrote about how payers are using AI against hospitals. By September, BCBSA had put a number on its side of the argument: $942 million in excess inpatient costs over two years, driven mostly by secondary diagnoses with no matching change in care. In the same month, HCA's CFO said hospitals are "behind the payers" on claims AI, and CMS Administrator Dr. Mehmet Oz said AI will raise healthcare costs in the short term because it makes billing more efficient.
What it means for 2027. Expect payers to use the BCBSA study in contract negotiations and audits, and expect CMS to look harder at coding intensity. Hospitals that use AI coding tools will need to show that added diagnoses match the care in the chart.
Budget line: Fund a coding and documentation audit. Compare CC/MCC capture rates by DRG before and after your AI coding tools went live, and check whether the fastest-growing secondary diagnoses have a treatment, order or monitoring plan behind them. Fund denial analytics too, so you can see where payer AI is hitting you. If you have the scale, fund a pilot for digital data exchange with one or two major payers, the approach HCA has used with five of its payers for nearly two years.
2. AI got cheaper to start and more expensive to keep
What happened. In March, I wrote about the AI ROI gap. In August, a UPMC and KLAS report found that 25 of 27 health system leaders had third-party AI running and only 11 had a place to test it. I also wrote about how cheap AI work can be to run (two weeks of work in four hours, for seventy-five dollars). Then Mayo Clinic, which has 128 clinical AI tools in use, said the cost of maintaining them is far higher than anyone expected. Health systems also started building kill switches and downtime plans for AI.
What it means for 2027. The first-year cost of an AI tool is getting easier to predict. The cost in year two and year three, when someone has to monitor the model, retrain it and check its accuracy, is where budgets fall short.
Budget line: Fund AI maintenance as its own line. For every AI tool already in production, budget for monitoring, accuracy audits and the staff time to review results. For any AI tool that has become the only way certain work gets done, fund enough trained people to cover an outage. Cut AI tools that can't show measurable value; negotiate exit clauses in renewals so you can.
3. Epic moved into almost every part of the revenue cycle
What happened. In March, I wrote that AI is moving inside the EHR. In May, I asked whether RCM vendors will survive, and in July I looked at where investors are placing their bets. By August, Epic had announced autonomous coding, a patient payment platform called Savvy, an AI agent platform and Rangers, its program to place Epic-employed staff permanently inside customers' IT departments. In September, CIOs said Epic was releasing more AI tools than they could test, and Epic paused most of its development for about six weeks to focus on security.
What it means for 2027. For Epic organizations, many third-party revenue cycle tools will compete with an Epic feature that is "good enough" and already in the license. Some third-party tools will still be worth keeping. Others will become duplicate spending.
Budget line: Fund a vendor review before renewals. List every third-party revenue cycle tool, what it costs, what it delivers in measurable results and whether Epic's roadmap overlaps it. Protect tools that can show results in your own data. Cut tools that duplicate what you already license. Decide in writing how long you'll wait for a promised Epic feature before you buy something that works now.
4. Prior authorization became public, and soon electronic
What happened. In April, I wrote that prior auth transparency is here: on March 31, payers had to publish their prior authorization denial rates for the first time. Electronic prior authorization requirements are scheduled to take effect January 1, 2027, and the AMA has asked CMS not to delay them. Aetna began bundling cancer prior authorizations into one request.
What it means for 2027. You now have public data on how your payers handle prior authorization, and electronic submission will start replacing fax and portal work for payers that are ready. Neither change helps much without trained staff and a baseline to measure against.
Budget line: Fund electronic prior auth training and measurement. Pay for staff training, payer and EHR connectivity, and a way to track turnaround time, staff touches and authorization-related denials. Pull your payers' published denial data and bring it to your next contract negotiation.
5. The coverage cliff got a date
What happened. I wrote in June that efficiency won't offset structural revenue loss, and in July about the hospitals already under financial strain. Medicaid work requirements start in 2027, and I wrote in September that your registration desk just became an eligibility office. CMS proposed a provider tax overhaul and a cap on state-directed payments worth a combined $1 trillion or more over a decade. Texas hospitals went two and a half weeks without authorization for nearly $12 billion in supplemental Medicaid funding. CMS canceled ACA coverage for 760,000 people.
What it means for 2027. Expect more self-pay patients, more patients who delay care until they need the ED, and more uncompensated care. Hospitals that rely on supplemental Medicaid payments face additional risk.
Budget line: Fund patient access and financial counseling. Staff for eligibility screening, Medicaid re-enrollment help and marketplace enrollment support. Model your self-pay and bad debt reserves on 2027 coverage assumptions, not 2026 actuals. Model supplemental Medicaid payments separately from base rates, and identify which service lines depend on them.
6. The revenue cycle workforce got squeezed from both ends
What happened. In April, I wrote about the workforce trap: cutting jobs and outsourcing without oversight. In August, I wrote that we're hiring for the technology and firing for the technology in the same fiscal year. Epic introduced an AI tool that does junior analyst work in minutes, which led health system leaders to ask how anyone becomes a senior analyst if the junior work disappears. Meanwhile, ChristianaCare used a Department of Labor grant to train employees for cybersecurity careers, graduated 87 of them and has hired three so far.
What it means for 2027. The roles revenue cycle needs most, like people who oversee automation, analyze denials and help patients with financial questions, are hard to hire from outside. Organizations that cut experienced staff before building those roles will lose knowledge they can't easily replace.
Budget line: Protect the people who understand the work, and fund the roles that oversee AI. Before cutting revenue cycle positions because of automation, fund automation oversight roles and cross-training for experienced coders, billers and access staff. Look for grant funding to pay for apprenticeships, as ChristianaCare did.
7. Governance and security became budget items
What happened. In April, I wrote that AI governance is coming to healthcare. In May, I wrote about the AI lawsuits, and in July about state AI laws filling the federal gap. In September, 26 state attorneys general asked Congress for federal AI safety rules, and I asked whether patients should know when AI calls about their bill. On security, I wrote that AI cyberattacks are a CFO problem. Sens. Warner and Wyden reintroduced a bill to set minimum cybersecurity standards for providers, payers and clearinghouses, with $1.3 billion to help hospitals comply.
What it means for 2027. More states will regulate how AI is used and disclosed, and federal cybersecurity standards may follow. Revenue cycle uses a lot of AI that touches patients directly, from billing calls to chatbots, so it will be part of those rules.
Budget line: Fund an AI inventory, a disclosure policy and a real downtime cost. Keep an inventory of every AI tool in the revenue cycle, who owns it and what it touches. Make sure revenue cycle has a seat on the AI governance committee. Set a policy on when patients are told AI is involved in billing. For security, calculate what a day of downtime would cost in delayed cash, and use that figure to size your security and continuity budget.
2027 budget questions for your finance committee
Bring these to your next budget review:
Do we know how our coding patterns changed after we started using AI coding tools, and can we defend the change in an audit?
Does every AI tool in production have a maintenance budget and a named owner?
Which AI tools have become the only way certain work gets done, and do we have trained people to cover an outage?
Which third-party revenue cycle tools overlap with features in our EHR license, and what do they deliver that the EHR doesn't?
Are we ready for electronic prior authorization on January 1, and have we measured our current turnaround times?
Have we pulled our payers' published prior authorization denial data for contract negotiations?
How are we modeling 2027 self-pay and uncompensated care, and does that model include Medicaid work requirements and ACA coverage losses?
How much of our revenue depends on supplemental Medicaid payments, and which service lines rely on it?
Are we cutting revenue cycle staff before we've built the roles that oversee automation?
What would one day of downtime cost us in delayed cash, and does our security budget reflect that?
Frequently asked questions
What were the biggest revenue cycle trends of 2026?
The seven biggest shifts were the public fight between payers and hospitals over AI coding, the rising cost of maintaining AI, Epic's expansion across the revenue cycle, prior authorization transparency and the coming electronic prior authorization requirements, Medicaid and ACA coverage losses, pressure on the revenue cycle workforce, and new attention to AI governance and cybersecurity.
What should hospitals prioritize in their 2027 revenue cycle budget?
Priorities include coding and documentation audits, a maintenance budget for AI tools already in use, a review of third-party vendors that overlap with the EHR, electronic prior authorization training, more patient access and financial counseling staff, roles that oversee automation, and a downtime cost estimate to size security spending.
How will Medicaid work requirements affect hospital revenue in 2027?
Work requirements begin in 2027 and are expected to reduce Medicaid coverage for some patients. Hospitals should expect more self-pay accounts and uncompensated care, and should plan for more eligibility screening and enrollment help at registration.
Why is AI raising healthcare costs?
AI coding tools can identify more billable diagnoses. Payers argue this raises costs without changes in care, while hospitals say it captures care that was previously under-documented. CMS's administrator has said AI will be inflationary in the short term.
Building your 2027 budget? These free resources can help:
For the full picture of where revenue cycle is headed by the end of the decade, read RCM 2030: Strategy and Survival for Revenue Cycle Leaders and the companion guides on operations, policy and regulation, and workforce modernization.
I cover what this means for health systems every Sunday in RCM 2030 Weekly. Subscribe on LinkedIn.
