Ike, I think there may be a broader Centene story hiding behind the workforce reductions.
Centene says it is building a more disciplined, technology-enabled operating model, but investors and employees have almost no visibility into what that actually means. The central question is whether Centene is building durable internal AI capabilities or using AI as the narrative wrapper for a large-scale reduction in employees, transfer of work to contractors and offshore vendors, and permanent hollowing-out of internal expertise.
A few things appear worth connecting:
Centene now projects roughly $315 million to $365 million in 2026 severance costs tied to enterprise optimization, yet only $18 million had been recognized through June 30. That jibes with the bulk of the workforce impact is still be ahead; what’s material here is that the company has not disclosed how many positions will ultimately disappear, how many employees accepted voluntary separation offers, which functions are being targeted, or its expected end-state headcount. Does anyone even know current headcount vs. expected headcount post-VSPs/layoffs??
I think the analysts missed some big opportunities to get to the heart of what really matters. Good luck with your article.
Also, very germane to the discussion are there are reports that Cognizant has secured a multiyear Centene technology-services agreement worth more than $500 million and potentially approaching $1 billion. Centene did not meaningfully address that reported agreement on its earnings call. That makes it important to determine whether laid-off employees’ work is being eliminated, automated or simply transferred to a vendor workforce. The distinction matters because all three may initially appear as lower SG&A, but they have very different long-term implications for capability, control and operating risk.
The AI piece of the puzzle also deserves scrutiny. Management regularly references digitization, automation and AI, but has not disclosed how much of projected cost savings comes from actual production automation, how much comes from headcount reduction and how much comes from outsourcing. There are few public metrics showing which workflows have been automated, the baseline labor involved, realized savings, error rates, member outcomes or the number of human decisions actually removed.
That creates a potentially important short-term vs. long-term story:
In the short term, replacing experienced employees with contractors or offshore vendor teams can improve reported expense ratios, convert fixed labor into variable spend and accelerate layoffs.
Over the longer term, Centene could lose institutional knowledge, internal technical ownership, bargaining power with vendors and the ability to govern AI systems independently. It may end up appearing more “digital” while becoming increasingly dependent on outside firms to understand and operate its own platforms.
There are already other signs that Centene may be moving risk away from the enterprise while presenting the change as innovation. The company placed more than 700 Indiana employees into an ICHRA model for 2026, using part of its own workforce to test a benefits product it is actively trying to sell. Centene describes this as increased employee choice, but it would be worth examining how employee premiums, deductibles, provider access, subsidies and out-of-pocket exposure compare with the previous group plan and whether the pilot is intended for wider rollout. See “Eating our own dog food.”
Someone else already mentioned this, but super important. There is a huge disconnect in Medicare. Management talks about improving Stars performance and rebuilding Medicare margins, but KFF found that only 6% of Centene’s 938,000 Medicare Advantage enrollees are in plans receiving quality bonus payments in 2026, compared with 68% across the industry. That is a major current economic disadvantage that has received remarkably little attention from analysts. It raises the question of whether near-term margin restoration is coming primarily from genuine quality improvement or from portfolio cuts, benefit reductions, settlements, pricing, workforce reductions and other cost actions.
Some questions that might open the story:
• What is Centene’s projected employee headcount when enterprise optimization is complete?
• How many employees have accepted voluntary separation offers, and how many involuntary layoffs are expected?
• Which functions are being transferred to Cognizant, Accenture or other vendors?
• Are affected Centene employees training vendor replacements before departure?
• What percentage of projected savings is attributed separately to AI, outsourcing, workforce reduction and traditional process improvement?
• How many contractors or vendor FTEs are being added while employees are being removed?
• Who owns the models, workflows, data pipelines and institutional knowledge after the work moves?
• What production metrics substantiate Centene’s AI claims with things like labor hours removed, cycle-time changes, accuracy, appeals, complaints and member outcomes?
• Is the Indiana ICHRA employee pilot expected to expand nationwide in 2027?
• How do employee costs and provider access under ICHRA compare with Centene’s previous group coverage?
• Why has management emphasized future Stars improvement without discussing that only 6% of current MA enrollment receives 2026 quality bonuses?
The broader thesis is:
Centene may be redesigning one of the largest healthcare operating models in America, but outsiders cannot currently tell whether its improving margins represent genuine AI-enabled productivity, conventional outsourcing, benefit-risk transfer, or the permanent reduction of internal capabilities.
Each could lift earnings in the short term. They do not create the same company (or the same risks) over the long term.
Good luck with the story.