Beyond "the cloud is safe": how zero trust security and data privacy protect people data
Moving people data to the cloud does not make it secure by itself; zero trust security, as described by NIST and CISA, protects it by verifying every user and device for every request, granting the least access needed, and assuming a breach is possible, while privacy practices limit what is collected in the first place.
HR data includes pay, performance, health and identity information, which makes it a target and a privacy obligation. Cloud providers secure their infrastructure, but how access is configured, who can see what, and what data is kept remain the customer's responsibility.

Why "the cloud is safe" is not enough
Cloud services operate under shared responsibility: the provider secures the platform, the customer secures accounts, permissions, configuration and data. Many incidents come from stolen credentials or overly broad permissions, not from the provider's infrastructure.
Zero trust in brief
NIST Special Publication 800-207 defines zero trust as moving defenses from network perimeters to users, assets and resources, with no implicit trust granted based on network location. CISA's Zero Trust Maturity Model organizes the work into five pillars: identity, devices, networks, applications and workloads, and data.
| Principle | In practice for HR data |
|---|---|
| Verify explicitly | Single sign-on with multi-factor authentication for every HR tool |
| Least privilege | Managers see their own teams; pay visible only to roles that need it |
| Assume breach | Encryption, audit logs, alerts on unusual access |
| Protect the data itself | Classify sensitive fields and restrict exports |
Privacy by design
- Collect less. Request only the fields a tool needs.
- Aggregate. Show group results with minimum group sizes where individual detail is not needed.
- Limit retention. Delete what is no longer needed.
- Be transparent with employees about what is collected and why.
The NIST Privacy Framework offers a structured approach to identifying and managing privacy risk alongside security.
Questions to ask any HR technology vendor
- Which fields do you read from our systems, and can we limit them?
- Do you support single sign-on and enforce multi-factor authentication?
- How are roles and permissions configured, and can pay be restricted?
- Is data encrypted in transit and at rest, and are access logs available?
- How is data deleted when we leave?
Organization Viewer's directory sync documents exactly which fields it reads and which it never requests, such as phone numbers and demographic fields. BlissPoint's Vendor Due Diligence project reviews vendors against questions like these.
Designing roles for HR data
| Role | Typically sees |
|---|---|
| Employee | Their own record and the directory |
| Manager | Their own team's records, not pay outside their team |
| HR partner | Records for their client groups |
| Compensation | Pay data, with export logged |
| Analyst | Aggregated or de-identified data where possible |
Review role assignments regularly and remove access promptly when people change jobs or leave.
Plan for an incident
Assume a breach is possible. Know which systems hold sensitive people data, who to call, how to revoke access quickly and how to notify affected people and regulators if required. Audit logs make it possible to answer what was accessed; keep them long enough to be useful.
Key takeaways
- Cloud providers secure their platforms; customers remain responsible for access, configuration and data.
- Zero trust verifies every user and request and grants least privilege, as NIST SP 800-207 describes.
- CISA's maturity model organizes the work into identity, devices, networks, applications and data.
- Privacy by design limits what is collected, shown and kept in the first place.
Questions
What is zero trust?
A security approach that grants no implicit trust based on network location and verifies every user, device and request, as defined in NIST SP 800-207.
Is cloud HR software secure?
The provider secures its platform; the customer must configure access, permissions and data handling securely.
What HR data is most sensitive?
Pay, health, identity numbers, performance and demographic data generally need the strictest controls.
Where do I start with zero trust?
Identity: single sign-on with multi-factor authentication and least-privilege roles. CISA's maturity model describes later stages.
Is multi-factor authentication enough?
It is a strong first step, but zero trust also needs least-privilege roles, monitoring, encryption and data controls.
Does zero trust apply to small companies?
Yes. The principles scale down: single sign-on with multi-factor authentication, least-privilege roles and limited data collection are achievable at any size.
How often should HR system access be reviewed?
At least quarterly for privileged roles such as compensation and administrators, and immediately whenever someone changes jobs or leaves. Automated removal tied to the HRIS reduces the chance of forgotten accounts.
How do we limit what HR vendors can see?
Grant integrations the narrowest scopes available, send only the fields each tool needs, and review vendor access when contracts renew.
What about data on employees' personal devices?
Zero trust treats device health as part of each access decision. Decide whether personal devices may reach sensitive HR data and under what controls.