TL;DR: Healthcare breach costs reached $7.42 million per incident for the 14th consecutive year, while stolen credentials stayed the top initial access vector and unpatchable medical devices kept widening exposure, according to Elisity’s recap of IBM, Verizon, and HIMSS findings. The governing assumption is no longer that entry can always be prevented, but that identity scope, segmentation, and lifecycle controls must limit how far an attacker can move once inside.
At a glance
What this is: This recap argues that healthcare security has shifted from perimeter defence to blast-radius control, with identity-based microsegmentation and lifecycle governance now central to limiting damage from stolen credentials, legacy devices, and AI-enabled identities.
Why it matters: For IAM, PAM, and NHI teams, the practical question is no longer only who can get in, but which identities, devices, and agents can reach anything useful after entry.
By the numbers:
- As covered in the recap, 60 percent of health systems say they cannot adequately protect the unpatchable, agentless medical devices already on their networks.
- The recap says 52 percent report they have no continuous monitoring for lateral movement or segmentation failures.
- The IBM Cost of a Data Breach Report 2025 put the global average breach cost at 4.44 million dollars.
- Healthcare again led every other sector at 7.42 million dollars per incident, according to the IBM Cost of a Data Breach Report 2025 cited by Elisity.
Context
Healthcare identity programmes now face a structural mismatch: attackers are not relying on noisy front-door attacks so much as stolen credentials, weakly monitored internal movement, and assets that cannot be patched on normal schedules. In that environment, the core security problem is blast radius, meaning how far an attacker can move once an identity or device is compromised.
The article frames identity-based microsegmentation as the architectural response because it ties policy to identity, not just network location. That matters in healthcare, where legacy medical devices, vendor access paths, service accounts, and AI agents all sit inside the same operational estate but should not share the same reach.
Key questions
Q: What breaks when healthcare networks still rely on flat internal trust?
A: Flat internal trust fails when a stolen credential or compromised device can move laterally without crossing meaningful barriers. In healthcare, that turns one foothold into a broad operational incident because legacy systems, vendor access, and critical clinical assets often sit too close together. The control failure is not just weak perimeter defence, but the absence of contained reach after entry.
Q: Why do legacy medical devices increase breach impact in hospitals?
A: Legacy devices often cannot support modern patching, strong authentication, or continuous monitoring, so they remain reachable long after vulnerabilities are known. That creates a durable attack path inside the clinical network. The risk is amplified when those devices share access paths with higher-value systems, because lateral movement becomes easier than remediation.
Q: How do security teams know if identity-based segmentation is actually working?
A: Teams should test whether a compromised or simulated compromised host can reach anything beyond the minimum required set of services. If lateral movement, management-plane access, or service-to-service discovery still succeeds, segmentation is too coarse. Effective segmentation shows up as denied pathways, reduced reachable surface, and a clear separation between user, management, and critical workloads.
Q: Should organisations treat AI agents and service accounts differently in SaaS governance?
A: They should govern them with the same relationship-aware discipline but not the same assumptions. Service accounts usually have more stable ownership and scope, while AI agents can chain access across workflows and integrations in ways that change during execution. That means both need inventory, lifecycle control, and blast-radius analysis, but agent behaviour deserves closer runtime scrutiny.
Technical breakdown
Why identity-based microsegmentation limits lateral movement
Identity-based microsegmentation constrains communication paths by binding policy to who or what the workload is, rather than to an IP range or flat subnet. In practice, that means a compromised credential does not automatically grant broad east-west movement. The model is especially relevant in healthcare because many assets are legacy, operationally sensitive, or impossible to patch quickly. By forcing every connection through explicit policy, segmentation converts a likely enterprise-wide event into a narrower, observable incident.
Practical implication: define network reach by identity and role, then deny implicit east-west trust.
How legacy medical devices change the access-control model
Legacy medical devices often cannot support modern authentication, multi-factor enforcement, or frequent patching, so they become durable footholds for attackers who already have credentials or an internal path. The article shows why compensating controls matter more than endpoint perfection in these environments. Privileged access gateways, dedicated VLANs, allow-lists, and logging are not optional extras when the device itself cannot be hardened to the same standard as a modern server or laptop.
Practical implication: wrap unpatchable devices in compensating access controls instead of assuming endpoint remediation will arrive on schedule.
Why AI agents and service accounts belong in the blast-radius model
The article extends the same containment logic to non-human identities, including AI agents, service accounts, and third-party connections. These identities can hold permissions, call APIs, and move data at machine speed, which means a weakly scoped account can amplify impact faster than a human operator can intervene. The governance lesson is that identity inventory and lifecycle control must include actors that execute work, not only people who log in.
Practical implication: inventory AI agents and service accounts as first-class identities with explicit scope and monitoring.
Threat narrative
Attacker objective: The attacker aims to turn one compromised identity or device into broad internal reach, reaching critical healthcare assets without tripping effective containment.
- Entry begins when attackers obtain stolen identities rather than breaking through the perimeter, which gives them a legitimate-looking foothold inside the environment.
- Escalation follows as they use that access to reach legacy devices and poorly monitored internal segments that cannot be patched or directly defended.
- Impact comes when lateral movement is left unconstrained, allowing the attacker to move from one reachable system to another until valuable assets are exposed.
Breaches seen in the wild
- Salt Typhoon telecom intrusions 2025: Salt Typhoon breached US telecoms mainly with stolen logins, then harvested SNMP strings and TACACS/RADIUS keys to spread and persist for years.
- Sisense breach 2024: A credential in Sisense's GitLab reportedly opened S3 buckets of customer tokens, passwords and certificates; CISA urged a full reset.
Read and download The State of NHI & AI Agent Breach Report 2026, covering 200+ breaches impacting Non-Human Identities including AI Agents.
NHI Mgmt Group analysis
Blast-radius control is now the right governing objective for healthcare identity security: Perimeter success no longer tells you whether the environment can survive compromise. In hospitals, the hard problem is limiting what a stolen identity can reach after initial access, especially when legacy systems and vendor paths remain unavoidable. The implication is that identity governance has to be judged by containment, not just authentication rates.
Legacy assets create a governance exception that cannot be solved by patch cadence alone: The article makes clear that many healthcare devices cannot be patched or instrumented like normal endpoints. That means the prevailing governance assumption that every asset can eventually be brought to a common hardening standard does not hold. Practitioners must treat unpatchable devices as a distinct containment class, not as temporary outliers.
Non-human identities now belong in the same containment model as human accounts: AI agents, service accounts, and third-party connections extend the blast radius when they are not inventoried, scoped, and monitored with the same discipline as employee access. This is the same identity problem expressed at machine speed, which means lifecycle governance has to span people, workloads, and agents together. The practitioner conclusion is to govern reach, not just login.
Identity-based microsegmentation is becoming the practical expression of Zero Trust in clinical environments: The article repeatedly shows that flat trust assumptions fail once an attacker is inside. Microsegmentation converts that insight into a control plane that can survive legacy constraints and operational reality. For healthcare security leaders, that means segmentation strategy is now a business continuity decision as much as a security design choice.
What this signals
Identity-based microsegmentation is becoming the practical control for healthcare blast-radius reduction: The article shows why containment now matters more than perimeter success when stolen credentials and unpatchable devices coexist in the same estate. Healthcare teams should expect segmentation projects to shift from network engineering tasks into identity governance programmes, because the policy boundary is increasingly defined by who or what is allowed to talk.
Non-human identities expand the attack surface in the same way as poorly governed human access: Service accounts, AI agents, and vendor connections all widen the reachable set if their permissions are not inventoried and constrained. That means lifecycle control and reach control need to be designed together, not treated as separate workstreams.
Healthcare leaders should expect containment metrics to become board-relevant: As breach costs stay high and legacy assets remain unavoidable, leaders will be asked not only whether controls exist, but how much damage a single compromised identity can do before detection and isolation kick in.
For practitioners
- Map blast-radius dependencies Identify which clinical systems, vendor pathways, service accounts, and AI agents can reach high-value assets if one identity is compromised. Use that map to set containment priorities before redesigning controls.
- Segregate unpatchable medical devices Place legacy devices into dedicated network segments with explicit allow-lists, controlled administration paths, and monitoring that compensates for the lack of direct hardening options.
- Remove implicit vendor reach Force third-party access through controlled privileged gateways and disallow direct remote administration paths that bypass segmentation and accountability.
- Inventory non-human identities by exposure Catalogue service accounts, AI agents, and workload identities by the systems they can touch, then classify them by blast-radius impact rather than by owner team alone.
- Tie logging to containment decisions Align monitoring with the segment boundaries that matter most so that lateral movement, segmentation failure, and anomalous access paths are visible before impact spreads.
Key takeaways
- The article argues that healthcare security now has to be measured by how much an attacker can reach after entry, not only by whether entry is blocked.
- Legacy medical devices, stolen credentials, and unmanaged non-human identities combine to make lateral movement the central operational risk.
- Identity-based microsegmentation and lifecycle governance are the controls that convert a broad compromise into a contained one.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
OWASP Non-Human Identity Top 10 and MITRE ATT&CK address the attack and risk surface, while NIST CSF 2.0 sets the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP Non-Human Identity Top 10 | NHI-05 — Overprivileged NHI | The recap centres on limiting what compromised identities can reach inside healthcare networks. |
| NHI-08 — Environment Isolation | Microsegmentation and dedicated device zones are the article's core containment mechanism. | |
| NHI-10 — Human Use of NHI | The article stresses vendor and operator pathways that should not bypass identity governance. | |
| Recommendation — Reduce standing reach for non-human identities and confine each credential to the smallest viable segment. Isolate legacy devices and vendor pathways into separate trust zones with explicit policy boundaries. Prevent human-driven shortcuts into NHI pathways and force administrative actions through governed access routes. | ||
| NIST CSF 2.0 | PR.AA-05 — Access Permissions, Entitlements and Authorizations | The piece argues for identity-scoped reach and explicit authorization boundaries. |
| PR.DS-01 — Data-at-Rest Protection | Damage limitation is about constraining access to critical assets when compromise occurs. | |
| Recommendation — Review entitlements so each identity only reaches systems required for its role and segment. Protect high-value clinical data with segmentation and access boundaries that limit downstream exposure. | ||
| MITRE ATT&CK | TA0006;TA0008 — Credential Access; Lateral Movement | The article explicitly describes stolen identities followed by internal movement. |
| Recommendation — Map stolen-credential and lateral-movement scenarios to these tactics and prioritise detection inside the network. | ||
Key terms
- Blast Radius: The potential scope of damage if a specific credential or identity is compromised. Identities with broad permissions have a larger blast radius and represent a higher priority for least-privilege enforcement and security controls.
- Identity-based Microsegmentation: A segmentation approach that uses identity, context, and policy to decide whether a connection should be allowed inside a network zone. In OT, it helps reduce lateral movement without relying only on IP addresses or broad subnet rules.
- Non-Human Identity (NHI): A digital identity assigned to a non-human entity such as a software application, service account, API key, bot, machine, or AI agent that enables it to authenticate and interact with systems without direct human involvement. NHIs now outnumber human identities in most enterprises by 25 to 50 times.
- Compensating Control: A compensating control is a measure that reduces risk when the ideal fix, such as immediate patching or redesign, is not possible. In OT, compensating controls often include session recording, access restriction, and tighter monitoring. They do not eliminate the underlying issue, but they narrow exposure until safer remediation can happen.
Deepen your knowledge
NHI governance, agentic AI identity, and machine identity lifecycle are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are building or maturing an IAM or NHI governance programme, it is worth exploring.
Published by the NHIMG editorial team on July 10, 2026.
Updated on October 11, 2026.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org