AI Usage Disclosure

Effective Date: August 9, 2026

1. Scope of This Disclosure

This AI Usage Disclosure applies to all commercial services provided by Eira OS, including infrastructure analysis, performance optimization, security assessments, compliance preparation, AI platform deployment, and ongoing monitoring. It is incorporated by reference into the Eira OS Terms of Service. This disclosure does not apply to consumer products, which are governed by separate terms and disclosures published on the respective product page.

2. How AI Is Used

Eira OS uses artificial intelligence, including large language models and machine learning tools, as part of its service delivery. AI is used to:

• Analyze infrastructure configurations, network architecture, and service deployments;
• Map compliance controls across multiple regulatory frameworks simultaneously;
• Identify potential security vulnerabilities, misconfigurations, and gaps;
• Generate draft documentation, including policies, procedures, and remediation roadmaps;
• Accelerate data analysis that would be impractical to complete manually within the Membership timeframe;
• Produce discovery knowledge bases compiling environment details into structured formats.

AI is used as a tool to accelerate and deepen analysis. It is not used as an autonomous decision-maker. No Deliverable is produced solely by AI without human review.

3. What AI Does

As part of Eira OS's expert-led, AI-amplified service model, AI performs the following functions:

1. Pattern Recognition: Identifies patterns, anomalies, and control gaps across large volumes of configuration data, policy documents, and system outputs.
2. Cross-Referencing: Maps controls across multiple compliance frameworks (e.g., NIST 800-171, CMMC, ISO 27001, SOC 2) to identify overlapping requirements and gaps.
3. Draft Generation: Produces draft text for policies, procedures, reports, and remediation recommendations based on analysis of Member Data.
4. Knowledge Compilation: Organizes discovered information into structured knowledge bases in requested formats (PDF, Word, HTML, Markdown).
5. Accelerated Analysis: Processes and analyzes data at a speed and scale that exceeds manual capability, enabling deeper analysis within Membership timeframes.

4. What AI Does Not Do

Specifically, Eira OS's AI does NOT:

• Make compliance determinations: AI identifies potential gaps and control mappings. A named expert validates every finding. A certified assessor performs any required certification.
• Issue certifications: Eira OS is not a C3PAO and does not employ a Qualified Assessor. AI cannot and does not issue any certification of any kind.
• Replace human expertise: AI amplifies human capability but does not replace human judgment. The expert-led model requires a named person to stand behind every finding.
• Operate autonomously: AI does not take actions, make changes, or interact with Member systems without explicit human direction. All AI operations are initiated and reviewed by Eira OS personnel.
• Make automated decisions about individuals: AI is not used to make decisions that produce legal effects or significantly affect individuals. See Section 9.
• Train on Member Data: Eira OS does not use Member Data to train, fine-tune, or improve AI models. Member Data is used solely for performing the Services during the Membership.

5. Human Validation Requirement

Every AI-assisted output must pass through a human validation process before it becomes a Deliverable. This process consists of:

1. Expert Review: A named Eira OS expert reviews every AI-generated finding for accuracy, completeness, and relevance.
2. Contextual Validation: The expert validates findings against the Member's actual environment, requirements, and business context.
3. Correction: AI errors are corrected before inclusion in Deliverables. Erroneous AI output is excluded.
4. Sign-Off: The named expert approves the final Deliverable. When a Member or auditor asks who reviewed a finding, there is a name. Not a dashboard.

6. AI and Your Data

Member Data provided to Eira OS for analysis may be processed by AI systems as part of Service delivery. The following applies:

• Where data is processed: Member Data is processed on Eira OS's sovereign, on-premise infrastructure. No Member Data is transmitted to third-party cloud AI services unless explicitly agreed in writing.
• How data is used: Member Data is used solely for performing the Services during the Membership. It is not used to train, fine-tune, or improve AI models.
• What happens after the Membership: Member Data is purged no later than thirty (30) days after the Membership concludes, unless a different timeline is specified in the Membership.
• Data ownership: The Member retains all ownership of its data. Eira OS acquires no ownership rights. See Section 7 of the Terms of Service.
• No data off platform: By default, no Member Data leaves Eira OS's controlled infrastructure. Any exception requires written agreement.

7. Sovereign AI Infrastructure

Eira OS operates its AI systems on sovereign, on-premise infrastructure. This means:

• AI models run on hardware controlled by Eira OS, not on third-party cloud platforms;
• No third-party processor has access to Member Data during AI processing;
• AI processing does not depend on external API availability or third-party service terms;
• All AI processing occurs within the United States;
• Infrastructure-level environmental impact is minimized through on-premise compute, reducing data center water and electricity consumption compared to cloud-based AI alternatives.

8. Your Rights Regarding AI

As a commercial Member of Eira OS, you have the following rights regarding AI usage:

1. Right to Know: You have the right to know that AI is being used in your Membership. This disclosure satisfies that right.
2. Right to an Explanation: You may request a human-readable explanation of how AI was used in producing any Deliverable. Eira OS will provide a summary of the AI-assisted process and the expert validation applied.
3. Right to Human Review: Every Deliverable has been reviewed by a named human expert. You may request to speak with the expert responsible for validating your Deliverables.
4. Right to Correct: If you identify an error in an AI-assisted Deliverable, Eira OS will investigate, correct the error, and re-validate the affected findings at no additional cost during the Membership term.
5. Right to Opt Out of Specific AI Uses: If you have a specific objection to a particular AI application in your Membership (e.g., you do not want AI to draft policies, only to analyze configurations), Eira OS will accommodate reasonable requests. This may affect the scope and timeline of the Membership.
6. Right to Data Return: Upon request, Eira OS will return all Member Data provided for AI processing and purge it from our systems per the data handling terms in your Membership.

9. No Automated Decision-Making

Eira OS does not use AI to make automated decisions that produce legal effects or significantly affect individuals. Specifically:

• AI does not make hiring, firing, credit, insurance, or benefits decisions;
• AI does not make compliance certification or regulatory determination decisions;
• AI does not make access control or permission decisions on Member systems;
• AI does not make purchasing or contractual decisions on behalf of the Member;
• All decisions with legal, regulatory, or business consequences are made by a named human expert, not by AI.

This section applies to Eira OS's commercial services. Consumer products, when available, will have separate disclosures addressing any automated features.

10. Accuracy and Limitations

AI systems have inherent limitations. Eira OS acknowledges these limitations and mitigates them through the human validation process described in Section 5. Known limitations include:

• Hallucination: AI may generate plausible-sounding but incorrect content. The expert validation process is designed to catch and exclude these outputs.
• Context Gaps: AI may not fully understand the Member's specific business context, industry nuances, or operational constraints. The expert provides this context during validation.
• Knowledge Cutoff: AI models may not reflect the most recent regulatory changes, threat intelligence, or industry developments. The expert supplements AI analysis with current knowledge.
• Bias: AI may reflect biases in its training data. The expert validates findings against the Member's actual environment, not against generalized patterns.

Despite these limitations, Eira OS's expert-led, AI-amplified model is designed to produce Deliverables that are more thorough and more accurate than either AI alone or manual analysis alone. The combination of AI depth and human judgment is the core of our service model.

11. Changes to This Disclosure

Eira OS may update this AI Usage Disclosure from time to time as AI technology, applicable regulations, or our service model evolve. The effective date at the top of this page indicates when the current version was posted. Changes apply to new Memberships entered into after the effective date. Memberships already in progress are governed by the disclosure in effect at the time the Membership was initiated, unless the Membership expressly provides otherwise.

12. Questions

If you have questions about how AI is used in your Membership, or if you wish to exercise any of the rights described in Section 8, contact Eira OS through our contact page. We will respond within a reasonable timeframe based on the nature of your request.

Relationship to Terms of Service: This AI Usage Disclosure is incorporated by reference into the Eira OS Terms of Service. In the event of a conflict between this disclosure and the Terms of Service, the Terms of Service control. This disclosure does not create any warranty not expressly stated in the Terms of Service or the applicable Membership.