Compliance & Security
Security You Can Trust.
Compliance You Can Prove.
CavalryAI is designed to protect sensitive organizational information across the entire AI workflow—combining privacy by design, permission-aware access, enterprise-grade controls and transparent governance.
Our Security Pillars
Built on a Foundation of Trust
Five principles guide how we protect your data and earn your trust every day.
Security First
Security is embedded in architecture, development and operational decisions—not added after deployment.
Privacy by Design
Sensitive data stays within controlled boundaries, with minimization, isolation and retention practices built around policy.
Access Control
Permission-aware workflows help ensure people and AI services can access only the information they are authorized to use.
Transparency
Policies, controls, auditability and supporting documentation are accessible to customers and reviewers.
Continuous Improvement
Security testing, monitoring, remediation and control reviews evolve as threats, technology and obligations change.
Compliance & Assurance
Meeting the Standards Your Organization Requires
Our controls are mapped to the frameworks that legal, regulated and public-sector organizations rely on—described at the level our evidence supports.
ISO/IEC 27001
Aligned with ISO/IEC 27001 controls
Our information security practices follow the ISO/IEC 27001 control framework for confidentiality, integrity and availability.
SOC 2
Aligned with SOC 2 Trust Services Criteria
Controls are organized around the SOC 2 criteria of security, availability, processing integrity, confidentiality and privacy.
HIPAA
Designed to support HIPAA-regulated workflows
Safeguards are designed to help healthcare and related organizations protect health information in AI workflows.
IRS Publication 1075
Controls designed to support IRS Pub. 1075 requirements
Controls are designed to support organizations that handle Federal Tax Information under IRS Publication 1075.
Data Protection & Privacy
Your Data Remains Under Your Control
CavalryAI is designed so customer information is protected throughout ingestion, retrieval, model interaction, storage and output handling.
- Customer data is not used to train public foundation models unless explicitly authorized and contractually agreed.
- Information is encrypted in transit and at rest.
- Access follows your organization’s existing permissions and policies.
- Retention, deletion and hosting options are defined with you and documented in your agreement.
Enterprise-Grade Security Controls
Layered Controls Across the Platform
- Encryption in transit and at rest
- Multi-factor authentication (MFA)
- Role-based and permission-aware access
- Audit logging and administrative accountability
- Vulnerability scanning and remediation
- Independent penetration testing where applicable
- Security monitoring and alerting
- Incident response and disaster recovery
- Secure development and change management
- Backup, recovery and business continuity
Compliance by Design
Compliance Is Not an Afterthought
CavalryAI is designed so security, access policy and governance travel with the AI workflow. Rather than treating compliance as a final checklist, the platform helps organizations apply controls at the point where data is accessed, models are selected and answers are delivered.
Let’s Build Secure AI—Together.
Have a security, privacy or compliance question? We’ll help your team understand how CavalryAI can fit your requirements.
Request Security Documentation →