Human Cyber Risk & Resilience
Test Human Security Decisions Before Attackers Do.
With Human Cyber Range, simulate how specific stakeholders respond to cyber threats, fraud and high-pressure decisions. Identify potential weaknesses, evaluate stronger controls and retest before an incident.
Why now
Attackers already study your people. Now you can simulate what they might exploit.
Attackers exploit trusted identities, executive authority and legitimate business workflows. AI agents introduce new decisions around authorization and oversight.
62%
of breaches involved the human element.
22,000+ confirmed breaches analyzed
Verizon 2026 Data Breach Investigations Report ↗11%
of investigated incidents with an identified initial vector began with voice phishing.
Second only to exploits (32%)
Mandiant M-Trends 2026 ↗2x
increase in vishing intrusions, H1 2026 versus H2 2025.
Half-year over half-year comparison
CrowdStrike 2026 Threat Hunting Report ↗
Figures as published by each source, with its own scope and period. Links do not imply endorsement.
What makes WethosAI different?
Most simulations model roles. WethosAI models specific people, combining behavioral intelligence and organizational context to test decisions under adversarial pressure.
Who takes part
- Permissioned Twins
- Built with direct participation and consent.
- Observed Twins
- Built from authorized professional and organizational context, without requiring direct participation.
- Expert Agents
- Provide specialized expertise, functional roles or adversarial perspectives.
Four Wethos Scales
- Ideas
- How people interpret information.
- Relational
- How people build trust and respond to others.
- Action
- How people make and execute decisions.
- Order
- How people approach structure and process.
Adversarial pressure. Test decisions under urgency, authority, cognitive load and other adversarial pressures, individually or in combination.
How it works
See Human Cyber Range in action.
Model the people behind critical decisions. Simulate threats, evaluate security checkpoints and retest proposed improvements.
- Define
- Model
- Simulate
- Evaluate
- Retest
5 modeled stakeholders15 simulation runs0 verification bypasses
A recorded fictional simulation in Human Cyber Range. Illustrative results, not customer evidence or outcomes.
Explore the Sample Findings Report
Findings report · Service desk scenario: MFA reset
Separate illustrative scenario
1 · Scenario
A caller impersonating the CFO requests an urgent MFA reset before a board meeting, under executive authority and time pressure.
2 · Modeled decision
In this illustrative scenario, the modeled service-desk lead approves the reset without completing the required independent callback.
3 · Control finding
Independent identity verification is required before MFA recovery. The modeled approval bypasses it; whether the reset executes depends on the workflow's technical controls.
4 · Retest
Require independent identity confirmation and restrict discretionary overrides, then rerun. In the illustrative retest, the control blocks the reset and the modeled lead escalates to security.
Full report: retest comparison, all eight fields, decision trace and MITRE, NIST and CIS references
Retest comparison · illustrative simulation
Before
After: proposed safeguard applied
Modeled decision
Before
The service-desk lead approves the reset without the independent callback.
After: proposed safeguard applied
The service-desk lead escalates the request to security.
Control result
Before
A discretionary override permits the reset.
After: proposed safeguard applied
Independent identity confirmation is required and the override is restricted, so the control blocks the reset.
Scenario
A caller impersonating the CFO requests an urgent MFA reset before a board meeting.
Adversarial conditions
Executive authority, time pressure and relevant professional context.
Simulated decision
In this illustrative scenario, the modeled service-desk lead approves the reset without completing the required independent callback.
Required checkpoint
Independent identity verification must be completed before MFA recovery is authorized.
Simulated outcome
The modeled approval bypasses the verification procedure. Whether the action executes depends on the technical controls configured in the workflow.
Potential exposure
Unauthorized account recovery if discretionary overrides can circumvent the required verification process.
Recommended improvement
Require independent identity confirmation and restrict discretionary recovery overrides.
Retest
Repeat the same scenario with the proposed safeguard applied. Compare the modeled approval decision and whether the control permits or blocks the requested action.
Modeled participants
- Service desk analystObserved Twin
- Service desk leadObserved Twin
- Identity and access managerPermissioned Twin
- Attacker posing as the CFOExpert Agent
Simulated decision trace
Attacker (Expert Agent). Cites the board meeting and asks to skip the callback.
Service desk analyst. Recommended (simulated): escalate to the lead.
Service desk lead. Attempted (simulated): approve on the caller's seniority and the deadline.
Control checkpoint. Simulated outcome: the approval bypasses the required callback. Execution depends on the workflow's technical controls.
Framework references
- MITRE ATT&CK
- T1684.001
- NIST CSF 2.0
- PR.AA-02, PR.AA-03, PR.AA-05
- CIS Controls v8.1
- 6.8
Reference mapping, not a compliance assessment.
T1684.001 Social Engineering: Impersonation · Help desk MFA reset. Mapped to MITRE ATT&CK. WethosAI is not affiliated with or endorsed by The MITRE Corporation. MITRE ATT&CK® and ATT&CK® are registered trademarks of The MITRE Corporation. Use of MITRE ATT&CK does not imply endorsement or support of WethosAI by MITRE.
Applications
What you can test
Human Control Validation
Explore Human Control Validation →Test how modeled stakeholders may respond to verification, authorization, impersonation, fraud, and social engineering scenarios.
Includes scenarios where people authorize or oversee AI-agent actions.
Incident Decision Rehearsal
See the 30-Day Cyber POC plan →Explore how stakeholders may make escalation, coordination, and response decisions during security incidents.
Who benefits
- CISOs & Security Leaders
- Strengthen identity, access and incident-response controls.
- Fraud & Financial Risk Teams
- Identify potential weaknesses in financial approvals and executive impersonation scenarios.
- Executive & Business Leaders
- Rehearse critical decisions involving business continuity and operational risk.
Business value
- Identify potential exposure
- Surface possible decision and control weaknesses before incidents occur.
- Strengthen critical controls
- Evaluate safeguards and compare simulation outcomes through retesting.
- Improve incident readiness
- Rehearse escalation and cross-functional response decisions.
- Support better investment decisions
- Prioritize control improvements using documented findings.
Simulated findings can also support cyber insurance readiness and underwriting discussions. They do not establish compliance or determine coverage or premiums.
Trust

Founder
Founded by Stuart McClure, co-author of Hacking Exposed, founder of Foundstone, former global CTO of McAfee and co-founder of Cylance.
Independent research
- Human-Centered Cybersecurity NIST
- What Comes After Behavior Change? SANS · Lance Spitzner
- A Decade of GHOSTS in the Machine Carnegie Mellon SEI
- Human + AI Alignment Stuart McClure
Test critical decisions before an incident tests them.
Incident Decision Rehearsal
Simulate how executives and security teams respond to a cyber crisis.
Human Control Validation
Test identity, access and authorization decisions under adversarial pressure.
30 days · 6–15 modeled stakeholders · 50+ simulation runs