Every PE firm with more than one healthcare platform eventually has to decide whether AI is something the fund drives or something each portfolio company figures out for itself. FTI Consulting's 2026 Private Equity AI Radar shows most funds have settled on a mix of both.
What Funds Are Doing
FTI asked respondents to rate each AI function from fully centralized at the fund to fully led by the portfolio company. The answers split by function:
- Mostly PortCo-led: delivery and implementation of AI use cases. 56% put it on the PortCo-led side of the scale.
- Mostly collaborative: shared AI delivery teams and AI/ML tooling and model platforms cluster in the middle.
- Where the fund leans in most: AI governance, risk and compliance, where 29% of respondents lean fund-led.
FTI's advice is that hybrid models "work best with clear operating models, defined roles and execution guardrails." Without those, hybrid means that nobody owns the work, and siloed ownership is already one of the top barriers to scaling AI in FTI's survey.
Why Healthcare Changes the Split
In most industries governance is something to coordinate. In healthcare it is a legal requirement. Every AI workflow that touches PHI raises HIPAA questions about where data lives, who can access it and whether a BAA is in place. A healthcare platform also has a third layer most portfolio companies lack: dozens of individual locations, each with its own staff and patients.
A Three-Layer Model for Healthcare Roll-Ups
| Layer | Owns | With Samara |
|---|---|---|
| Fund / sponsor | Value creation targets, KPI oversight, portfolio-wide standards | Comparable KPIs across platforms and portfolio companies |
| Platform (MSO / DSO) | Data, governance, compliance, the operating playbook | The Rollup Engine (AIP) unifies data and standardizes the playbook on infrastructure the platform owns |
| Location | Daily execution with patients | The Operations Engine (AIT) runs scheduling, intake, recall and communication at each site |
This keeps what has to be consistent in one place: data, compliance and the playbook. Execution stays close to patients. It matches what FTI's respondents are doing, with a clear owner at each layer.
Keep Governance Close to the Data
Centralizing governance is much easier when the data is actually in one place. On an Enterprise Sovereign AI OS, the platform's data runs on infrastructure it owns, either hosted and managed by Samara on a secure GCP instance or self-hosted by the client. Access, audit and PHI handling are defined once for the whole platform. They are not renegotiated for each tool at each location. See ownership and hosting models and HIPAA compliance.
Let the Fund See Across Platforms
FTI found that performance tracking and KPI oversight sit in the middle of the scale, shared between fund and PortCo. That only works if the KPIs mean the same thing everywhere. When every location runs on the same OS, the sponsor can compare fill rates, no-shows and new-patient bookings across platforms, and every new portfolio company starts from the proven playbook.
Next Step
Read about Samara for private equity, or book a demo to design the AI operating model for your platform.
Frequently Asked Questions
Do PE funds centralize AI or leave it to portfolio companies?
Mostly both. FTI Consulting's 2026 Private Equity AI Radar found delivery is largely PortCo-led (56%), while governance, risk and compliance is where funds lean in most.
What should be centralized in a healthcare roll-up?
Data, governance, HIPAA compliance and the operating playbook should sit at the platform level. Daily patient-facing execution should run at each location.
How does a fund compare AI results across platforms?
By running every location on the same operating system with KPIs baselined before go-live, so metrics like fill rate and no-show rate are comparable everywhere.