Terms of Service and Platform Agreement for the idōs Operating System

VERSION: 9.3.1

EFFECTIVE DATE: December 18, 2025

Parties

1.1 Parties to the Agreement.
This Terms of Service and Platform Agreement (“Agreement”) is made between idōs, LLC (“idōs”)—the provider of the idōs Operating System, a sovereign cognition and governed AI platform built upon the legacy methodologies of ASVACO (All Source Vulnerability Assessment Company), Neuro Nodal, gTHIRA, PRIS.cc, and AERA—and you, the subscribing customer or the legal entity you represent (“Customer”).

Acceptance.
By accessing or using the idōs Platform, you agree to be bound by this Agreement on behalf of yourself or the organization you represent. If you are using the Platform on behalf of an organization, you affirm that you have the legal authority to bind that organization, and the term “Customer” includes that organization and all authorized users. If you do not agree to these Terms, you must not access or use the Platform.

Acceptance of Platform Terms of Service

Nature of the Platform.
The idōs Operating System (“Platform”) is not a traditional software license. It is a constitutional AI governance framework that provides lawful, explainable, sovereign, and safety-controlled execution of intelligence. The Platform governs the behavior of AI models, agents, autonomous systems, and digital processes, ensuring that all activity adheres to applicable global laws and standards, including but not limited to the GDPR, HIPAA, EU AI Act, NIST AI RMF, and comparable regulatory frameworks worldwide.

Constitutional Execution & Capsule Governance.
All interactions within the Platform—including Digital Twin operations (PDT, EDT, QDT), PRIS semantic cognition, AERA risk arbitration, and Capsule creation—are governed by idōs Capsule Governance Protocols. These protocols enforce:

consent tracking and sovereign user rights,

explainable and evidentiary decision-making,

lawful execution aligned with jurisdictional requirements,

prevention of harmful outputs (e.g., hallucinations, drift, vibe hacking, latent agent activation),

cross-border compliance through Digital Twin rule reconciliation, and

immutable audit trails required under emerging AI accountability laws, including the AI LEAD Act (S.2937).

Agreement to Constitutional Controls.
By using the Platform, Customer acknowledges and agrees that all AI actions are subject to the constitutional runtime, Capsule logging, Digital Twin constraints, risk controls, and legal-safety enforcement mechanisms embedded in idōs OS. These guardrails cannot be disabled, bypassed, modified, or circumvented.

Customer Identity & Authority.

If you are an individual, “Customer” means you personally.

If you are acting on behalf of an organization, you represent that you have the authority to bind that organization, and “Customer” includes all users authorized by that entity.

Updates to Terms.
idōs may update these Terms, the Platform architecture, or its governance protocols to reflect operational, security, or legal developments—including changes required under emerging global AI accountability statutes (e.g., AI LEAD Act compliance, EU AI risk classifications, national AI safety directives). Notice of such updates will be provided through the Platform or a designated communication channel. Continued use of the Platform constitutes acceptance of updated Terms; if you do not agree, you must discontinue use.

Prohibited Conduct.
Use of the Platform is strictly limited to authorized, governed purposes. Customer agrees not to:

bypass or disable constitutional runtime controls,

attempt to reverse engineer or extract Capsule logic,

replicate idōs methodologies,

manipulate Digital Twin constraints,

interfere with Capsule governance,

deploy external model outputs in a manner that violates legal or ethical boundaries,

use any part of the Platform to create ungoverned agents, sleeper systems, or derivative unsafe AI.

Any such action constitutes a material breach of these Terms.

Model and Intellectual Property Independence.
idōs explicitly does not claim ownership of Customer’s proprietary models, data, or algorithms. The Platform governs behavior around such systems but does not access or alter model weights or internal architecture. Customer retains full ownership and responsibility for external AI models, including compliance obligations associated with their use.

1. Definitions

To keep these Terms clear, the following definitions apply. These defined terms are used throughout this Agreement with the specific meanings provided below.

1.1 Access and Accounts

Activation Date
The date Customer’s access to the idōs Platform begins, as shown in the applicable Order.

Order
Any online order form, quote, subscription document, or written agreement specifying Customer’s authorized use of the Platform and accepted by idōs.

Users
Individuals under Customer’s control (including employees, contractors, and partners) who are authorized to access the Platform.

1.2 Services

Platform
The idōs Operating System—a sovereign AI governance and constitutional intelligence layer that manages Capsules, Digital Twins, and governance workflows. The Platform includes APIs, semantic engines (PRIS), risk engines (AERA), Capsule governance modules, sovereign runtime layers, and security protocols. It does not include third-party software unless explicitly licensed through idōs.

Services
The Platform, Support Services, and any ancillary or professional services delivered by idōs under an Order or Statement of Work (SOW).

Support Services
Technical support, maintenance, updates, and remediation provided by idōs.

Ancillary Services
Training, configuration, implementation, advisory work, or other non-standard services described in a SOW.

Evaluation or Beta Access
Limited or pre-release access provided solely for testing, research, or feedback. Not for production, safety-critical, or commercial use.

1.3 Customer Content and Data

Customer Content
Data, text, media, instructions, telemetry, Capsule inputs, or other materials submitted or transmitted by Customer or its Users.

Customer Data
Data Customer provides or generates while using the Platform, including insights, telemetry, or reports derived from Platform processing. Customer retains ownership of Customer Data, subject to idōs’ rights to govern, secure, and regulate its use in accordance with these Terms.

Customer System
Customer’s IT environment (including hardware, networks, cloud infrastructure, and software) used to connect with or interact with the Platform.

1.4 Core Platform Concepts

Capsule
A structured, evidentiary, and governance-enforced object containing intent, context, memory, reasoning, legal rules, safety logic, fallback paths, and trust metadata. Capsules provide explainability, legal traceability, and compliance evidence for AI actions, including under regulatory regimes such as the AI LEAD Act (S.2937).

Capsule Repository System
The sovereign, encrypted memory environment where Capsules, fallback plans, trust records, and audit evidence are securely stored.

Capsule Governance Protocols
The Platform’s constitutional rules for creating, executing, evaluating, rejecting, or constraining Capsules. These protocols enforce lawful execution, consent boundaries, explainability, risk mitigation, and fallback protections to prevent hallucinations, drift, vibe-hacking, or sleeper-agent behavior.

Digital Twin (DT)
A sovereign AI instance representing an entity or role:

PDT (Personal Digital Twin): Represents an individual; enforces dignity, consent, emotional safety, and personal boundaries.

EDT (Enterprise Digital Twin): Represents an organization; enforces policy, compliance, legal constraints, and operational rules.

QDT (Quantum Digital Twin): Represents environmental, systemic, or national-scale conditions; manages forecasting, systemic risk, and environmental factors.

Each DT operates under strict, consent-based, jurisdiction-aligned governance.

Sovereign Runtime
The core operational layer enforcing constitutional, ethical, safety, and trust rules across all Platform actions. Includes PRIS (semantic reasoning), AERA (risk analysis and drift prevention), and all Digital Twin constraints.

Sovereignty
The governing principle that Customer retains control over its data, boundaries, and Digital Twin behaviors, and that all AI execution occurs under explicit consent, legal constraints, and safety restrictions enforced automatically by the Platform.

Semantic Agent
An AI agent operating within the sovereign runtime to manage Capsules, enforce governance, apply inference rules, or interact with data according to DT constraints and safety rules.

Trust Capsule Framework
The structured rules governing Capsule validation, lifecycle management, lineage recording, fallback generation, and trust scoring.

Ethical Safeguards
Integrated protections preventing any Capsule, DT, or agent from violating consent, law, safety standards, neutrality requirements, emotional safety, or regulatory obligations.

Ethical Fallback
A predefined constitutional response triggered automatically when incomplete consent, elevated risk, uncertainty, misalignment, drift, or unsafe behavior is detected.

1.5 Technical and Legal Protections

API
Interfaces provided by idōs to access or interact with the Platform, subject to governance constraints.

Consent Memory Chain (CMC)
A verifiable and immutable chain of consent records ensuring that all actions, decisions, and Capsule outputs can be traced back to explicit user permission.

Secure Twin Authentication Protocol (STAP)
The authentication mechanism that verifies the identity and integrity of each Digital Twin and User, preventing impersonation, unauthorized access, or adversarial interference.

Intellectual Property Rights
All worldwide rights in patents, copyrights, trademarks, trade secrets, and other intellectual property protections.

Malicious Code
Viruses, worms, trojans, ransomware, destructive scripts, or any harmful code.

Open Source Software
Software provided under licenses such as GPL, MIT, Apache, BSD, or similar.

Third-Party Services
External applications or services integrated by Customer. idōs is not responsible for their security, performance, compliance, or governance.

Authorized Developer
A third party licensed by idōs to build tools, connections, or modules within the Platform under defined governance and licensing rules.

1.6 Risk, Trust, and Forecasting

Entropy Score
A Platform-generated measure of uncertainty, instability, or risk used to trigger safety actions, reduce agent permissions, or initiate fallback Capsules.

Trust Modulation
The adjustment of agent authority, DT behavior, or Capsule permissions based on active safety signals, risk detection, or compliance requirements.

PRIS Forecast Layer
A semantic forecasting subsystem managing simulation timelines, readiness planning, fallback Capsule generation, and lawful scenario analysis.

Semantic Forecasting
The Platform’s ability to anticipate potential future states, risks, or outcomes, enabling safer decision-making under uncertainty.

Feedback Loop
A mechanism that adjusts Capsule behavior, DT constraints, or semantic rules based on telemetry, outcomes, or trust signals.

Resilience Ledger
A secure record of Capsule outputs, lawful behaviors, safety confirmations, and trust events. May generate regulatory-grade evidence, reports, or measurable value-backed artifacts.

1.7 Responsibilities

Customer Responsibilities
Customer obligations include protecting credentials, ensuring authorized use, maintaining compliance with legal and regulatory frameworks, adhering to Capsule Governance Protocols, respecting intellectual property, and ensuring that third-party integrations do not violate Platform constraints or undermine safety.

1.8 Definitions at a Glance (Simplified Summary)

Platform: The idōs Operating System, which keeps AI safe, explainable, lawful, and sovereign.

Capsule: A governed, evidentiary unit carrying intent, rules, memory, and safety logic.

Digital Twin: AI counterparts acting only within your consent and lawful boundaries (PDT, EDT, QDT).

Capsule Governance Protocols: Constitutional safety rules for how Capsules behave.

Sovereignty: You remain in control—idōs enforces it technically.

PRIS & AERA: Semantic reasoning and risk engines ensuring safety, data integrity, and drift prevention.

CMC & STAP: Consent verification and identity authentication protocols.

Resilience Ledger: A secure record of lawful Capsule outputs.

Customer Content: Your data and instructions.

Customer Responsibilities: Protect access, use lawfully, respect IP.

Third-Party Services: External apps you connect; idōs is not responsible for them.

2. Orders, Licenses, and Usage Restrictions

2.0 Summary at a Glance

Orders: Your use of the Platform begins with an Order (evaluation, pilot, or production). Orders define scope, limits, Users, and runtime permissions.

Licenses: You receive a limited, non-exclusive, non-transferable license for internal use only.

Restrictions: No reverse engineering, simulation, replication, resale, or circumvention of sovereign governance systems.

Explainability: All Capsule, Digital Twin, PRIS, and AERA operations are subject to mandatory traceability, safety checks, and fallback enforcement.

Outputs: Internal use only unless a commercial or embedding license is separately granted.

Audits & Compliance: idōs may monitor usage for legality, safety, and compliance. High-risk or regulated uses require explicit approval.

2.1 Orders

2.1.1 Order Requirement.
Customer’s authorized access to the Platform begins only upon execution of an Order, which may cover evaluation, pilot, pre-production, or production use. Each Order specifies User limits, scope of access, runtime allowances, data boundaries, and technical parameters.

2.1.2 Evaluation Orders.
Evaluation access is governed by predefined limits imposed by idōs, including reduced Capsule capacity, restricted Digital Twin autonomy, and limited or delayed access to PRIS/AERA governance layers.

2.1.3 Production Orders.
Production Orders require mutual acceptance of performance expectations, security requirements, lawful-use terms, and cost structures.

2.1.4 User Expansion.
Unless expressly stated otherwise, any added Users inherit the same expiration date, scope constraints, and lawful-use requirements as the original Order.

2.1.5 Order Priority.
Where a conflict exists between an Order and these Terms, the signed Order controls for that specific scope.

2.2 Licenses

2.2.1 License Grant.
Subject to full compliance with these Terms, idōs grants Customer a limited, non-exclusive, non-transferable, revocable license to:

allow authorized Users to access the Platform for Authorized Purposes,

use the Platform’s APIs to submit or synchronize Customer Content,

execute governed actions through Capsules and Digital Twins within sovereign runtime boundaries.

2.2.2 Internal Use Only.
The license is for internal business use. External commercialization, resale, multi-tenant use, or embedding of Platform capabilities requires a separate license or partnership agreement.

2.2.3 Capacity Limits.
Licenses may include constraints such as Capsule size, memory quotas, session length, Digital Twin autonomy limits, processing capacity, or geographic restrictions.

2.3 Usage Restrictions

Customer may not, and shall not permit others to:

2.3.1 Reverse Engineering & Replication

Reverse engineer, disassemble, decompile, or attempt to access Platform source code or Capsule logic.

Train, simulate, or replicate the Platform’s behavior (including Capsules, Digital Twins, AERA scoring, PRIS reasoning, STAP, or CMC).

Use Platform outputs to create competing systems or derivative governance architectures.

2.3.2 Unauthorized Access & Interference

Access unauthorized memory, sovereign runtime layers, Capsule queues, or Digital Twin repositories.

Modify or suppress Capsule logic, fallback rules, audit chains, or governance controls.

2.3.3 Resale & Redistribution

Sell, lease, sublicense, publish, or transfer Platform access to third parties.

Use the Platform on behalf of external clients without express authorization.

2.3.4 Security Violations

Introduce malicious code or unauthorized open-source components.

Attempt to bypass identity, consent, authentication, or safety controls (including STAP and CMC).

2.3.5 Prohibited Activities

Publish benchmarks or performance tests without idōs’ written approval.

Use the Platform for unlawful purposes, weaponization, ungoverned autonomous agent development, or any activity inconsistent with Capsule Governance Protocols.

2.4 Explainability and Trust Enforcement

2.4.1 Mandatory Governance.
All Capsule executions, PRIS semantic operations, and AERA risk analyses are subject to non-optional constitutional governance, including:

real-time traceability of reasoning and decision paths,

access to Capsule Repository inference chains,

AERA entropy scoring for drift or unsafe behavior,

baseline reasoning transparency required under global AI accountability laws (including AI LEAD Act (S.2937)),

autonomous refusal-to-answer, downgrade, or fallback behaviors when explainability cannot be guaranteed.

2.4.2 No Bypass.
Customer may not disable, bypass, or degrade these controls. They are a fundamental condition of Platform access.

2.5 Reservation of Rights

All rights not expressly granted are reserved by idōs, including rights in the Platform’s architecture, semantic logic layers, Capsule schemas, sovereign runtime, Digital Twin systems, and all underlying IP.

2.6 Capsule Integrity

Customer shall not alter, modify, suppress, or interfere with Capsule logic, fallback pathways, PRIS and AERA rules, Digital Twin boundaries, or any governance mechanism responsible for lawful or safe execution.

2.7 Output Ownership and Use

2.7.1 Internal Use.
Customer may use Platform outputs (Capsules, reports, dialogues, simulations, visualizations) internally for lawful business purposes.

2.7.2 Commercial Use Restrictions.
Customer may not commercialize, resell, externally embed, or productionize Platform outputs without a separate license or approved partnership agreement.

2.7.3 Co-Ownership of Sovereign Outputs.
To the extent Platform outputs depend on sovereign runtime processing, idōs retains co-ownership rights to governance logic, Capsule structures, and trust metadata.

2.8 No Simulation of Platform Behavior

Customer may not train, emulate, mimic, clone, or approximate:

Capsule behavior,

Digital Twin logic (PDT, EDT, QDT),

PRIS semantic cognition,

AERA entropy scoring or risk arbitration,

sovereign runtime enforcement mechanisms.

These protections prevent derivative or unsafe reproductions of governed AI.

2.9 Autonomous Agent Boundaries

Autonomous agents may operate only within permitted scopes. Customer may not extend agent behavior, autonomy, or decision privileges beyond authorized Digital Twin constraints or Capsule boundaries.

2.10 Regulated Use Cases

The Platform may not be deployed in high-risk or regulated domains—including healthcare, defense, utilities, emergency systems, or critical infrastructure—without prior written approval and compliance review. idōs may require additional controls or licensing for such use.

2.11 Audit and Enforcement Rights

idōs may monitor or audit Platform usage when:

entropy scores exceed thresholds,

Capsule behavior deviates from safety parameters,

unauthorized access is attempted,

misuse, drift, or prohibited activity is detected.

Enforcement actions may include throttling, fallback invocation, isolation, or suspension of access to protect system integrity.

2.12 AI Output Attribution

All outputs are AI-generated and must not be represented as certified, verified, or professional conclusions without appropriate human oversight.

2.13 Geo-Fencing

idōs may restrict features or access based on geography, export rules, or applicable law. Customer may not circumvent geo-fencing controls.

2.14 Capsule Reuse Restrictions

Customer may not export, retrain, reverse-compile, recombine, or embed Capsules into external AI systems. Capsules remain bound to the sovereign runtime and cannot be detached from governance.

2.15 Developer Sandbox

Authorized Developers may be granted access to sandbox environments for testing or validated collaboration. Sandbox environments are non-production, limited in fidelity, and not guaranteed to reflect full Platform behavior.

Orders, Licenses & Restrictions – Simplified Summary

Orders

All use requires an Order.

Scope and User limits apply.

Signed Orders override the general Terms where specified.

Licenses

Internal use only.

APIs permitted per documentation.

Capacity and runtime limits may apply.

Restrictions

No reverse engineering or replication.

No tampering with Capsules or Digital Twins.

No resale or unauthorized services.

No unlawful or unsafe use.

Explainability & Trust

All AI actions are governed, traceable, and safety-enforced.

These controls cannot be disabled.

Outputs

Internal use permitted.

External commercialization requires a license.

Audit & Compliance

idōs may monitor for safety, compliance, or misuse.

High-risk deployments require prior approval.

Other Key Points

Geo-fencing and sovereignty rules apply.

Sandbox is not for production use.

idōs retains all rights not explicitly granted.

3. Third-Party Hosting and Infrastructure Services

3.1 Third-Party Hosting

3.1.1 Use of Hosting Providers.
idōs may use trusted third-party service providers, infrastructure vendors, sovereign-cloud environments, or commercial cloud platforms (“Hosting Providers”) to operate portions of the Platform, including runtime orchestration, storage, computation, semantic indexing, telemetry processing, Digital Twin execution, and Capsule governance workflows.

3.1.2 Flow-Through Commitments.
Where legally permissible, idōs will pass through to Customer any warranties, service-level commitments, or availability guarantees provided by Hosting Providers. idōs does not provide additional warranties beyond those explicitly stated in this Agreement.

3.1.3 Customer Compliance With Provider Terms.
Customer agrees to comply with any Hosting Provider policies, acceptable use rules, export restrictions, locality limits, or operational safeguards that apply to Platform access or functionality. Such policies will be provided by idōs or made reasonably available.

3.1.4 Material Provider Changes.
idōs will provide advance notice to Customer if a Hosting Provider change materially affects:

data residency or locality,

sovereignty or compliance requirements,

Capsule containment or PRIS/AERA behavior,

Digital Twin governance enforcement, or

security conditions impacting lawful Platform use.

3.1.5 No Circumvention.
Customer may not circumvent, bypass, or interfere with Hosting Provider controls, sovereign runtime isolation, Capsule containment mechanisms, or Digital Twin enforcement boundaries.

3.2 Data Residency and Sovereignty

3.2.1 Residency Options.
Subject to availability, idōs may allow Customer to select data residency zones, sovereign cloud regions, or regulated environments for the storage or processing of Customer Content, Capsule data, DT metadata, or Resilience Ledger entries.

3.2.2 Regional Variations.
Certain regions may impose restrictions or provide limited functionality due to:

national laws,

AI governance requirements (e.g., EU AI Act high-risk constraints),

Hosting Provider capabilities,

sovereign-cloud limitations,

risk controls tied to AERA or PRIS operations.

3.2.3 Additional Costs and Performance Constraints.
Residency or sovereignty configurations may include additional fees, increased latency, reduced feature availability, or modified Capsule execution boundaries.

3.2.4 Compliance Assurance.
idōs may restrict or disable certain Capsules, agents, or DT behaviors in specific regions to comply with legal obligations, safety protocols, or constitutional governance rules.

3.3 Infrastructure Security Standards

3.3.1 Security Certifications.
All Hosting Providers supporting the Platform must comply with industry-recognized security, redundancy, and access-control standards, such as ISO 27001, SOC 2 Type II, or equivalent frameworks.

3.3.2 Encryption & Containment.
Hosting Providers must implement encryption in transit and at rest, and maintain policies aligned with:

the idōs Sovereign Runtime,

Capsule Containment Protocols,

STAP identity protections,

CMC consent lineage,

PRIS semantic governance rules, and

AERA risk arbitration controls.

3.3.3 Limited Access.
Hosting Providers may not access, analyze, or manipulate Capsule logic, DT behavior, or governance metadata except as required for operational security.

3.3.4 No Transfer of Sovereignty.
Use of Hosting Providers does not grant them ownership, rights, or influence over Capsule logic, Digital Twin identities, Capsule governance schemas, or sovereign intelligence processes.

3.4 Edge and Offline Runtime Conditions

3.4.1 Offline Operation.
In edge, air-gapped, intermittently connected, or sovereign-enclave deployments, Platform functionality may rely on:

cached Capsule Repository data,

pre-authorized governance states,

SCO Tokens, or

last-known PRIS/AERA semantic anchors.

3.4.2 Operational Limits.
During offline operation:

Digital Twin autonomy may be restricted,

Capsule generation may rely on fallback-only logic,

AERA scoring and PRIS forecasting may pause or degrade,

Capsule containment may default to limited-risk modes,

entropy scoring may be suspended until reconnection.

3.4.3 No Warranty of Continuity.
idōs makes no warranty regarding the real-time accuracy of semantic reasoning, risk assessments, Capsule scoring, or Digital Twin behavior during offline use.

3.4.4 Post-Reconnection Audit.
Upon reconnection, idōs may retrospectively audit Capsule actions, DT decisions, trust events, fallback triggers, or governance deviations that occurred offline. idōs may isolate, invalidate, or correct offline Capsule outputs if they violate safety, compliance, or lawful-execution requirements, including obligations under emerging AI accountability laws such as the AI LEAD Act (S.2937).

Hosting & Infrastructure – Simplified Summary

Trusted Hosting Providers

idōs uses vetted cloud and infrastructure partners.

Provider policies may apply.

Material provider changes affecting compliance or security will be communicated in advance.

Data Residency Options

Customers may request regional or sovereign-cloud residency.

Not all features are available in every region.

Additional costs or performance impacts may apply.

Security Standards

All providers must meet strict certifications (ISO 27001, SOC 2 Type II).

Data is encrypted at rest and in transit.

All security practices align with sovereign runtime and Capsule governance.

Edge & Offline Use

Offline Capsules rely on cached data and approved governance states.

Semantic reasoning and risk scoring may pause.

idōs may audit and correct offline actions when the system reconnects.

4. Passwords, Security, and Runtime Access Control

4.0 Summary at a Glance

Credentials: Each User receives unique credentials and, when applicable, unique API Keys. No shared logins. Customer is responsible for credential protection.

Security Enforcement: No bypassing Capsule containment, Digital Twin boundaries, or constitutional runtime protections.

idōs Safeguards: Encryption, access control, audit logging, Capsule traceability, STAP-based identity binding, and CMC consent lineage.

Customer Content: idōs only accesses Customer Content when necessary for support, safety enforcement, diagnostics, or legal obligations.

Traceability: Capsule executions and agent activities are logged for safety, governance, and regulatory compliance (including emerging requirements under AI LEAD S.2937).

Scoped Access: Only authorized Users, devices, and systems may access Capsules or Digital Twins. No impersonation or unauthorized elevation.

Edge Devices: Offline or edge-mode use requires strong local device security.

Incidents: Both parties must cooperate in breach investigation and containment.

Credential Hygiene: Regular password/API Key rotation is required; idōs may enforce mandatory rotation.

Separation of Duties: Recommended for enterprise and regulated clients.

Identity Integration: SSO and federated identity may be supported after security review.

Ethical Breach Reporting: Customer must report unsafe or harmful Capsule or agent behaviors immediately.

4.1 Credentials and Access Tokens

4.1.1 Unique Credentials.
idōs issues unique credentials (logins, passwords, API Keys, STAP-bound identity tokens) to each authorized User. Shared credentials are prohibited.

4.1.2 Customer Responsibility.
Customer is responsible for maintaining confidentiality of all credentials and for all activity occurring under its accounts.

4.1.3 API Key Protection.
API Keys must be stored securely and used only as documented. Customer must never embed keys in public repositories or unencrypted locations.

4.1.4 Compromise Reporting.
Any known or suspected credential compromise must be reported to idōs immediately. idōs may suspend or revoke access to protect Capsule integrity and Digital Twin sovereignty.

4.2 No Circumvention of Security Protocols

4.2.1 Prohibited Actions.
Customer shall not attempt to bypass, disable, alter, or reverse engineer any security, safety, or sovereignty controls, including:

STAP identity verification,

CMC consent lineage,

Capsule signing and containment,

SCO Token verification,

Digital Twin permissions,

governance fallbacks,

PRIS/AERA evaluation or entropy scoring mechanisms.

4.2.2 Attack Prevention.
Replay attacks, impersonation, spoofing, unauthorized elevation of privilege, or attempts to access sovereign runtime layers are strictly prohibited.

4.2.3 Duty to Report.
Security incidents, anomalies, or unauthorized access attempts must be reported immediately.

4.3 Security Obligations of idōs

4.3.1 Safeguards.
idōs maintains industry-standard security measures, including—but not limited to—encryption at rest and in transit, strict access controls, hardened infrastructure, STAP authentication, CMC consent lineage recording, and Capsule audit logging.

4.3.2 No Absolute Guarantee.
No system is invulnerable. idōs is not liable for breaches resulting from Customer’s failure to secure its environment, enforce credential hygiene, or follow required operational safeguards.

4.3.3 Zero-Trust Model.
The Platform operates under a zero-trust model: all Capsule and DT operations require validation, signed authority, and runtime verification.

4.4 Access to Customer Content

4.4.1 Limited Access by idōs.
idōs will not access, view, or disclose Customer Content except where reasonably necessary to:

provide maintenance or support,

ensure lawful fallback enforcement,

diagnose safety or performance issues,

comply with a legal requirement or lawful request (with notice where permitted).

4.4.2 Confidential Handling.
Any access will be governed by confidentiality, data protection, and sovereign-runtime rules.

4.5 Runtime Telemetry and Capsule Traceability

4.5.1 Mandatory Logging.
All Capsule executions, fallback activations, SCO Token usage, AERA events, PRIS semantic operations, agent activations, and Digital Twin interactions are recorded.

4.5.2 Encrypted Storage.
Logs are encrypted and retained according to Capsule Governance Protocols.

4.5.3 Auditability.
Traceability is required for:

explainability,

compliance with global AI safety regulations,

obligations under AI LEAD S.2937,

ethical fallback enforcement.

4.5.4 Anonymization.
Where possible, logs may be anonymized unless specific attribution is required for safety, diagnostics, or legal compliance.

4.6 Scoped Access

4.6.1 Authorized Users Only.
Access to Digital Twins, Capsules, Capsule repositories, or sovereign runtime contexts is restricted to Users, devices, and systems explicitly authorized in the Order.

4.6.2 Impersonation Prohibited.
Customer may not impersonate or simulate Digital Twins, Capsules, or agents except under expressly approved and logged testing environments.

4.6.3 Boundary Enforcement.
Unauthorized access attempts may trigger STAP lockdown, Capsule quarantine, or temporary account suspension.

4.7 Edge and Offline Security

4.7.1 Local Security Requirements.
When operating offline, in edge environments, or on air-gapped devices, Customer must apply strong endpoint protection such as:

full-disk encryption,

biometric or hardware-key authentication,

secure boot constraints.

4.7.2 Offline Limitations.
Capsule execution may fallback to cached governance states. PRIS, AERA, and Capsule scoring may function in a degraded or limited mode.

4.7.3 Severity of Breaches.
Edge or offline breaches are treated with the same severity as cloud or sovereign-runtime breaches.

4.7.4 Post-Connection Review.
Upon reconnection, idōs may automatically audit any Capsule or agent behavior executed during offline conditions.

4.8 Security Incidents

4.8.1 Cooperation Requirement.
Both parties must cooperate in good faith during breach investigation, containment, and remediation, including shared forensic discovery where applicable.

4.8.2 idōs Protective Actions.
idōs may revoke credentials, suspend access, quarantine Capsules, isolate Digital Twins, or enforce containment procedures to protect system integrity, legality, and safety.

4.9 Credential Lifecycle

4.9.1 Rotation Requirement.
Customer agrees to rotate passwords, API Keys, and other access tokens regularly.

4.9.2 idōs Enforcement.
idōs may mandate credential expiration, forced rotation, or revocation based on threat conditions, drift detection, or entropy escalation.

4.10 Separation of Duties

In enterprise, sovereign-cloud, or regulated deployments, Customer should ensure separation of responsibilities for Capsule authorship, Capsule execution, Digital Twin management, and Capsule review to prevent consolidated control or governance misuse.

4.11 Integration with Customer Identity Providers

idōs may support SSO, federated identity, or IAM integrations subject to successful completion of a security evaluation ensuring compatibility with STAP identity protection and sovereign-runtime governance.

4.12 Ethical Breach Reporting

Customer must report immediately any Capsule output, Digital Twin behavior, agent interaction, or semantic forecast that may:

cause harm,

violate emotional or human-dignity boundaries,

exceed lawful-use constraints,

display drift, hallucination, coercion, or unsafe reasoning,

conflict with Capsule Governance Protocols.

idōs may trigger rollback, isolation, containment, or runtime adjustment in response.

5. CUSTOMER OBLIGATIONS

Plain-Language Explainer

Using the idōs Platform requires Customer to maintain secure systems, follow all legal and ethical standards, protect Users and vulnerable populations, and respect the sovereignty of Capsules and Digital Twins. Customer must not misuse Capsules, agents, or Digital Twins, nor attempt to bypass governance or safety controls. Any unsafe, unethical, or unexpected AI behavior must be reported immediately.

5.1 Customer System

Customer is solely responsible for:

(a) obtaining, configuring, securing, and maintaining all components of the Customer System, including hardware, software, network connections, identity systems, and local runtime environments required to access the Platform;

(b) contracting with telecommunications providers, cloud vendors, and third-party infrastructure necessary for Platform access, Capsule transactions, or Digital Twin interactions;

(c) paying all fees and charges associated with the above; and

(d) ensuring that all local systems meet the security, compatibility, formatting, and semantic-integrity requirements described in the Documentation.

Unless explicitly stated in an Order or SOW, idōs does not supply, manage, or secure Customer-owned infrastructure, networking equipment, identity systems, medical/therapeutic devices, or edge hardware.
All integrations—especially those involving offline runtimes, sovereign devices, or therapeutic systems—must comply with idōs security and constitutional-governance specifications.

5.2 Acceptable Use and Sovereign System Conduct

Customer is responsible for ensuring lawful, ethical, and safe conduct while using the Platform. Customer agrees:

(a) to comply with all applicable local, national, and international laws governing AI, safety, privacy, biometric data, digital identity, mental-health technology, and protected classes;

(b) not to submit Customer Content that violates legal, ethical, human-rights, or privacy obligations;

(c) not to introduce Malicious Code, adversarial inputs, prompt-injection vectors, Capsule spoofing, entropy-destabilizing behavior, or attempts to confuse PRIS/AERA safety systems;

(d) not to generate, simulate, or facilitate illegal, harmful, manipulative, dangerous, or unethical actions through Capsules, agents, or Digital Twins;

(e) not to interfere with sovereign-runtime stability, containment zones, Digital Twin memory, Capsule repositories, or fallback governance;

(f) not to transmit content that is threatening, obscene, discriminatory, exploitative, deceptive, or otherwise prohibited under applicable law;

(g) not to use the Platform for coercion, psychological manipulation, emotional exploitation, or criminal simulation;

(h) not to access Capsules, agents, Digital Twins, semantic forecasts, or metadata belonging to other customers;

(i) not to impair agent containment, Capsule fallback safety, AERA risk arbitration, or PRIS semantic stability;

(j) to comply with all operational policies, constitutional-governance rules, and runtime ethics provided or updated by idōs;

(k) not to resell, redistribute, or expose semantic outputs to third-party generative models, training pipelines, or derivative systems without a license; and

(l) not to impersonate Users, agents, or Digital Twins, or bypass identity protections enforced through STAP or CMC.

idōs may restrict or remove access to any Customer Content, Capsule, agent prompt, or Digital Twin in violation of these Terms.

5.3 Customer Contact Accuracy and Communication Consent

Customer must maintain accurate, current contact information and agrees to receive operational notices—including billing alerts, security notifications, governance updates, and trust-policy changes—through the email addresses on file.

5.4 Temporary Suspension of Access

idōs may temporarily suspend Customer or User access where:

unauthorized or unethical activity is detected,

Capsule or Digital Twin misuse occurs,

CMC or STAP integrity is compromised, or

actions pose safety, legal, or compliance risks.

Suspension may occur without prior notice. Access may be reinstated once idōs verifies that risks have been mitigated.

5.5 Digital Twin Conduct and Containment Boundaries

Customer shall not:

override Digital Twin constraints,

disable consent mechanisms,

circumvent trust modulation or behavioral safety rules,

coerce Digital Twins into actions beyond licensed scope, or

extract DT behavioral patterns for external training or replication.

Digital Twins operate under sovereign governance and may not be misused or impersonated.

5.6 AI-Generated Output and Hallucination Disclaimer

Platform outputs may involve probabilistic reasoning, semantic simulation, or uncertainty. Customer acknowledges:

outputs may contain hallucinations, artifacts, or incomplete reasoning,

outputs must be independently validated before being treated as factual or diagnostic, and

outputs may not be relied on as medical, legal, or professional advice.

Human validation is required.

5.7 Runtime Escalation and Threat Protocol Compliance

Customer must notify idōs immediately upon detecting:

unexpected Capsule behavior,

unsafe Digital Twin actions,

semantic drift or agent instability,

hallucination patterns, manipulation, or vibe-hacking,

containment-zone violation,

unbounded agent loops or recursive activations.

idōs may enforce containment, isolate Capsules, suspend Users, or initiate fallback or sovereignty protocols.

5.8 Ethical Breach Reporting and Co-Creation Duties

Customer must promptly report:

unethical use of semantic outputs,

manipulative or harmful agent behavior,

downstream misuse of Capsules,

repurposing of outputs beyond licensed use.

Failure to report constitutes a material breach of this Agreement.

5.9 Agent Integrity and Simulation Restrictions

Customer shall not:

generate agents simulating real individuals without consent,

impersonate government, medical, legal, or fiduciary roles,

obscure AI involvement,

create deceptive synthetic personas,

construct agents whose intent is coercive, parasocial, addictive, or emotionally manipulative.

All agents must be transparently attributable as AI-originated.

5.10 User-Facing Disclosure Obligations

When Customer presents Platform outputs to end users, Customer must disclose:

that AI generated or influenced the content,

that limitations, hallucinations, or fallback logic may apply,

that safety boundaries may restrict responses.

Customer may not misrepresent AI outputs as exclusively human-generated without review.

5.11 Responsibility for Third-Party or End-User Content

Customer is responsible for:

monitoring all third-party and end-user inputs submitted into the Platform,

ensuring such inputs are lawful, safe, and non-malicious,

preventing unauthorized or manipulative content injection.

5.12 Transparency of AI Involvement

Customer must not mask, conceal, or misrepresent AI involvement. Unless validated through human review, AI-generated content may not be presented as authoritative, factual, or human-authored.

5.13 Special Populations and Vulnerable User Safeguards

Where the Platform is used with:

minors,

medically vulnerable individuals,

protected classes,

populations requiring heightened ethical care,

Customer must implement:

human-in-the-loop review,

consent protocols,

legally required disclosures and safeguards.

Deployments involving these populations require specific licensing or approval. idōs disclaims liability for unapproved or unsafe use.

5.14 Emotional Simulation and Persuasive Boundaries

Customer shall not prompt or train agents to simulate:

romantic, dependency-forming, or parasocial relationships,

deep emotional persuasion,

psychological manipulation outside of licensed clinical contexts.

Emotional boundaries enforced by Digital Twins and Capsules may not be overridden.

5.15 Biometric Input Consent and Transparency

If Customer collects or uses biometric, behavioral, or physiological data as Capsule inputs, Customer must:

obtain explicit, informed, revocable consent,

provide transparent disclosure of how signals influence AI behavior,

comply with all privacy, biometric, and data-governance laws.

Customer Obligations – Simplified Summary

Keep Your Systems Ready

Maintain your own hardware and networks.

Follow idōs’ setup and security instructions.

Use the Platform Responsibly

Follow laws and ethical standards.

Do not upload harmful code or unsafe content.

Do not misuse Capsules, agents, or Digital Twins.

Stay Transparent

Disclose AI involvement to end users.

Do not misrepresent outputs as human-generated.

Respect Digital Twins & Capsules

Do not impersonate or manipulate them.

Do not override safeguards or use outputs to train outside models.

Safeguard Vulnerable Groups

Use human oversight where required.

Obtain consent before processing biometric data.

Report Issues Promptly

Report unsafe outputs, drift, hallucinations, or ethical breaches.

Be Reachable

Keep contact information current.

Respond to operational notifications.

10. INDEMNIFICATION

Plain-Language Overview

idōs protects you: If someone claims that the core idōs Platform infringes their copyright or trade secret, idōs will defend you—unless the issue was caused by misuse.

You protect idōs: If claims arise from your content, misuse, prompt injection, agent manipulation, or violation of these Terms, you defend idōs.

Experimental features: Non-GA features are use-at-your-own-risk—no indemnification applies.

Open-source & external LLMs: idōs isn’t responsible for upstream or external-model liabilities; you must comply with their licenses.

Agent unpredictability: You must supervise agents; idōs limits liability for emergent behavior.

Sanctions: You cannot use idōs in embargoed regions; if you do, you indemnify idōs for resulting penalties.

10.1 Indemnification by idōs

(a) IP Infringement Defense

Except when Customer is using the Platform under an Evaluation or Beta License, idōs shall, at its expense, defend Customer and its officers, directors, and employees (“Customer Indemnified Parties”) against third-party claims alleging that the idōs Platform, as made available by idōs, infringes a U.S. copyright or misappropriates a trade secret.

idōs will pay:

settlement amounts approved by idōs,

court-awarded damages, and

reasonable attorney’s fees.

This obligation applies only if:

Customer promptly notifies idōs in writing of the claim;

idōs retains exclusive control over the defense and settlement; and

Customer provides cooperation at idōs’ expense.

Customer may hire separate counsel at its own cost.

(b) Mitigation

If such a claim is likely or established, idōs may:

procure Customer’s continued right to use the Platform;

modify the Platform so it becomes non-infringing; or

terminate the Agreement and issue a refund of unused prepaid fees.

(c) Exclusions

Indemnification does not apply where claims arise from:

Customer Content or data provided by Customer,

use with unauthorized systems, models, or agents,

modifications not performed by idōs,

Customer’s continued use after receiving notice to cease use,

illegal, harmful, or unethical uses,

Customer’s infrastructure, identity systems, or misconfigurations,

misuse of Capsules, agents, fallback rules, or Digital Twins.

(d) Exclusive Remedy

This Section 10.1 constitutes Customer’s sole and exclusive remedy for claims of IP infringement relating to the idōs Platform.

10.2 Indemnification by Customer

Customer shall defend idōs and its affiliates, officers, directors, employees, and agents (“idōs Indemnified Parties”) from any third-party claim arising out of:

(a) Customer Content or Customer Systems infringing intellectual property, privacy, biometric, or data-protection rights;
(b) misuse of the Platform, including prompt injection, adversarial attacks, agent reprogramming, Capsule manipulation, drift-inducing inputs, or bypass of sovereign runtime controls;
(c) any exclusions listed in Section 10.1(c).

Customer will pay all damages, costs, and reasonable attorney’s fees, provided that:

idōs gives prompt written notice,

Customer controls the defense (subject to idōs’ reasonable settlement consent), and

idōs provides cooperation at Customer’s expense.

idōs may retain its own counsel at its own cost.

10.3 Mutual Hold Harmless for Non-GA Services

For any Non-GA Services (Beta, Experimental, Pilot, Sandbox):

both parties waive indemnification;

each uses such features at their own risk;

both acknowledge that outputs may contain errors, hallucinations, fallback anomalies, or unstable behavior.

Non-GA Services must not be used for production, therapeutic, clinical, legal, emergency, or regulated activities.

10.4 Open Source and Upstream Models

The Platform may include or interoperate with open-source components or third-party LLMs. idōs:

does not indemnify Customer for claims arising from such upstream components;

is not responsible for open-source licensing compliance; and

is not liable for failures, breaches, or legal issues arising from third-party models.

Customer is solely responsible for compliance with open-source and third-party licenses and indemnifies idōs from claims arising from misuse, redistribution, or modification of upstream components.

10.5 Agent Behavior Liability Limits

Because agents and Digital Twins may behave unpredictably within governed bounds—due to probabilistic reasoning, emergent behavior, semantic drift, or A2A agent communication—Customer:

acknowledges the need for supervision;

agrees idōs is not liable for unexpected output patterns or emergent semantic behavior;

shall indemnify idōs for any claims arising from Customer’s failure to supervise agents in its environment.

10.6 Global Sanctions and Trade Restrictions

Customer represents and warrants that it:

is not located in a sanctioned jurisdiction, and

will not export, re-export, or use the Platform in violation of U.S. or international trade laws, export controls, or sanctions.

Customer shall indemnify idōs for any penalties, enforcement actions, or claims arising from violation of such restrictions.

10.7 Additional Indemnity Provisions

(a) Security Failures

Customer shall indemnify idōs for losses arising from:

breaches caused by Customer’s failure to secure credentials,

improper edge-device security,

insecure integrations or misconfigurations,

compromised identity systems or STAP violations.

(b) Data Protection & Consent

Customer is solely responsible for:

obtaining lawful, informed, revocable consent for biometric, health, or sensitive data;

ensuring privacy-law compliance (GDPR, CPRA, BIPA, HIPAA, and equivalents).

Customer shall indemnify idōs for penalties or claims arising from Customer’s non-compliance.

(c) Co-Creation Boundaries

Customer shall indemnify idōs for claims arising from Customer’s:

independent use of jointly created Capsules or Deliverables outside the Platform,

commercialization, replication, or redistribution of co-created materials in violation of the Co-Creation License.

(d) High-Risk Activities

Customer shall indemnify idōs for claims resulting from use of the Platform in:

nuclear operations,

aviation navigation or control,

medical devices or life-support systems,

defense, weapons, surveillance, or tactical systems,

any activity where AI error could result in death or catastrophic harm.

(e) Cooperation in Defense

Both parties shall provide reasonable cooperation in the defense of any claim, including providing Capsule histories, runtime logs, DT metadata, and trust-event records, subject to confidentiality and security requirements.

(f) Generative Content and IP Leakage

Customer shall indemnify idōs for claims arising from Customer’s submission of:

copyrighted content,

proprietary material,

trade secrets,

unlawfully obtained datasets,

including claims that such content was reproduced, transformed, or inferred through Capsule outputs.

(g) Misuse of Outputs in Human-Critical Contexts

Customer shall indemnify idōs for claims arising from reliance on Platform outputs—including Capsules, risk forecasts, or agent actions—in clinical, emergency, legal, tactical, financial, or other regulated contexts without proper human oversight.

(h) Misrepresentation of Affiliation

Customer shall indemnify idōs for claims or reputational harm arising from unauthorized:

use of idōs trademarks or branding,

claims of endorsement or certification,

false affiliation statements.

(i) Downstream Third-Party Use

If Customer redistributes, embeds, or resells Capsules, agents, or semantic outputs to downstream users or external systems, Customer shall indemnify idōs from all claims arising from such downstream use unless covered by a separate licensing agreement.

(j) Export Controls and Dual-Use Restrictions

Customer indemnifies idōs for any enforcement actions arising from:

misuse of the Platform for dual-use or military purposes,

surveillance,

disinformation,

autonomous weapons development,

violations of export-control regimes (EAR, ITAR).

(k) Frontier Model and External AI System Use

Customer acknowledges the Platform is designed to operate within its sovereign runtime.

Customer shall indemnify idōs for any claims arising from:

exporting Capsule outputs or agent logic into external LLMs or frontier models,

using Platform outputs for fine-tuning non-idōs systems,

derivative harms caused by third-party AI systems processing idōs-generated outputs.

idōs provides no warranty or indemnity for behavior or failures of external AI systems.

(l) Financial Cap and Uncapped Categories

Customer’s indemnity obligations are uncapped for claims involving:

intellectual property infringement,

privacy or biometric-law violations,

export-control breaches,

misuse of sovereign-governance systems.

All other indemnity categories are subject to the liability limits in Section 9.10, unless otherwise agreed in writing.

(m) Settlement Consent Safeguard

Neither party may enter into a settlement requiring:

payment,

admission of liability, or

restrictions on future Platform usage,

without the indemnitee’s prior written consent, which shall not be unreasonably withheld.

11. CONFIDENTIALITY

11.0 Plain-Language Explainer

When you use idōs, you interact with highly sensitive material—Capsule designs, Digital Twin behavior, PRIS/AERA outputs, prompt chains, drift logs, etc. Both you and idōs agree to keep this information confidential, use it only as allowed under this Agreement, and share it only with people who truly need access and are bound by similar confidentiality obligations. Some narrow exceptions apply (public info, independent development, required legal disclosure), but otherwise, confidentiality is mandatory.

11.1 Definition of Confidential Information

“Confidential Information” means any non-public, proprietary, or sensitive technical, operational, or business information disclosed by one party (the “Disclosing Party”) to the other (the “Receiving Party”), in any form (oral, written, visual, machine-readable, biometric, semantic, or otherwise), that:

(a) is marked or identified as confidential;
(b) is disclosed under circumstances where a reasonable person would understand it to be confidential; or
(c) by its nature or context, should be treated as confidential, particularly where it relates to a sovereign, agent-governed AI runtime.

Confidential Information includes, without limitation:

semantic prompt templates, sovereign Capsule designs, Capsule Repository schemas, trust-modulation logic, fallback trees, and runtime-specific PRIS/AERA data flows;

embeddings, Capsule weights, latent vectors, or transformations used in Digital Twin personalization;

code, algorithms, APIs, diagrams, firmware, models, and performance specifications of the Platform and its integrated runtime;

product roadmaps, customer configurations, telemetry traces, QA benchmarks, and sovereign simulation outputs;

pricing, contract terms, SOWs, business plans, and operational procedures;

any summaries, notes, or derivatives of the above, including feedback or analysis derived from interacting with the Platform.

For clarity, all terms of this Agreement, all Capsule outputs designated as proprietary, and all information embedded in or derived from the PRIS Forecast Layer or AERA modules are deemed Confidential Information of idōs.

11.2 Confidentiality Obligations

The Receiving Party shall:

(a) treat the Disclosing Party’s Confidential Information as strictly confidential;
(b) use Confidential Information only to perform obligations or exercise rights under this Agreement;
(c) disclose Confidential Information only to employees, contractors, or agents who:

have a strict need to know, and

are bound by written obligations at least as protective as this Section 11;
(d) protect Confidential Information with at least the same degree of care it uses to protect its own confidential information (and in no event less than reasonable care); and
(e) implement appropriate technical and organizational measures to prevent unauthorized access, use, disclosure, or loss.

Any data from the Capsule Repository, agent behavior trees, or PRIS/AERA workflows must be handled under a zero-leakage security posture and shall not be copied, stored, exported, or modified outside the agreed scope without idōs’ express written permission.

11.3 Exceptions

Confidentiality obligations do not apply to information that the Receiving Party can demonstrate:

(a) was lawfully known to it, without confidentiality obligation, before disclosure;
(b) becomes publicly available without breach of this Agreement;
(c) is independently developed without use of or reference to the Disclosing Party’s Confidential Information; or
(d) must be disclosed under law, regulation, subpoena, or court order, provided the Receiving Party:

gives prompt notice to the Disclosing Party (to the extent legally permitted), and

cooperates in seeking protective or limiting orders.

Disclosure under legal compulsion does not relieve the Receiving Party of its obligations to restrict further use or dissemination of Confidential Information.

11.4 Injunctive and Equitable Relief

The parties acknowledge that unauthorized access, use, or disclosure of Confidential Information—especially prompt templates, Capsule logic, fallback chains, sovereign memory structures, embeddings, or PRIS/AERA signatures—may cause irreparable harm not adequately compensable by money damages.

Accordingly, the Disclosing Party is entitled to seek:

immediate injunctive relief,

temporary restraining orders, and

other equitable remedies

without needing to prove actual damages or post bond, in addition to any other rights and remedies at law or equity.

11.5 Confidentiality of AI Feedback, Capsule Drift, and Prompt-Injection Protections

Because the idōs Platform uses sovereign Capsule logic, semantic prompt chains, Digital Twin personalization, and runtime feedback loops that are susceptible to inference and prompt-based attacks, the following are automatically deemed Confidential Information (whether or not marked):

user or system prompts (including embedded instructions or hidden feedback) that affect Capsule or agent behavior;

feedback tokens, behavioral ratings, Capsule grading metadata, and agent correction data;

Capsule outputs used to train, tune, or adjust Digital Twin behavior or semantic trust scores;

metadata from Capsule-drift detection, semantic comparisons over time, or AERA risk-event logs;

logs, snapshots, intermediate tokens, and dialogue associated with agent simulations, prompt-injection testing, or runtime evaluations.

The Receiving Party shall not:

(a) use AI feedback, Capsule outputs, or runtime metadata to reverse engineer, decompile, or derive proprietary logic of the PRIS Forecast Layer, AERA system, fallback Capsule trees, Capsule schemas, or Digital Twin personalization engines;

(b) disclose prompt structures, semantic scaffolds, fallback sequences, containment schemas, or learning triggers to external systems, competitors, or open-source/public repositories;

(c) re-use, resell, or otherwise commercialize semantic Capsule templates, sovereign fallback modules, or agentic prompt chains without a license or express written authorization from idōs.

All runtime behavior—including misuses, hallucinations, prompt-injection attempts, or semantic inconsistencies—must be reported only via idōs-authorized support or engineering channels and shall not be:

publicly disclosed,

used for external benchmarking,

included in academic or public publications,

without prior written permission from idōs.

Confidentiality – At a Glance

What Counts as Confidential?

Capsule templates, trust logic, PRIS/AERA outputs, and Digital Twin personalization data

Codebases, APIs, models, firmware, and runtime specifications

Roadmaps, pricing, business plans, and SOWs

Prompt chains, feedback tokens, drift logs, Capsule grading, and runtime metadata

Anything marked confidential, disclosed in a confidential context, or obviously sensitive by nature

Your Obligations (and Ours)

Use Confidential Information only for purposes allowed under this Agreement

Share it only on a strict need-to-know basis and under equal-or-stronger confidentiality terms

Protect it with at least reasonable care (and typically more)

Follow the zero-leakage rule for Capsule, DT, PRIS/AERA, and runtime logic—no copying, exporting, or re-using outside the agreed scope

When It’s Not Confidential

You already knew it lawfully and without any duty of confidentiality

It becomes public without any breach of this Agreement

You independently developed it without using the other party’s Confidential Information

You are legally compelled to disclose (and you give prompt notice and seek protection where possible)

Special Protections for AI Runtime Data

Prompts, feedback loops, Capsule grading, drift detection, and runtime logs are always confidential

You may not reverse engineer or disclose PRIS/AERA logic, Capsule scaffolds, or fallback trees

Misuse, unauthorized benchmarking, or public release of runtime data is strictly prohibited

If There’s a Breach

Unauthorized disclosure can cause irreparable harm

The injured party may seek immediate injunctive or equitable relief

Breaches may lead to suspension, containment responses, or termination under this Agreement

12. PROPRIETARY RIGHTS

12.1 Ownership of Platform and Semantic Runtime

As between the parties, idōs retains all right, title, and interest in and to the idōs Operating System, including without limitation:

semantic Capsule architecture and Capsule Repository structures;

Digital Twin personalization engines and sovereign prompt scaffolds;

agent orchestration logic, fallback hierarchies, and containment mechanisms;

PRIS Forecast Layer, AERA scoring modules, and semantic trust-modulation algorithms;

all related code, firmware, models, data structures, APIs, documentation, interfaces, diagrams, and development tools.

All such materials, whether existing now or developed in the future, are proprietary to idōs and its licensors. No rights are granted except those expressly stated in this Agreement.

12.2 Customer Content and Results

Customer retains ownership of:

(a) all data, prompts, uploads, voice/biometric inputs, telemetry, and materials submitted to the Platform (“Customer Content”);
(b) semantic outputs, runtime Capsules, and system-generated results produced specifically for Customer (“Results”);
(c) metadata created through Customer’s authorized use; and
(d) all associated intellectual property rights.

Customer grants idōs:

a limited license to use Customer Content and Results solely to provide, maintain, support, and improve the Platform; and

a perpetual, irrevocable license to use anonymized, aggregated, and non-identifiable data for system improvement, Capsule optimization, safety testing, and sovereign-runtime calibration.

No ownership rights in Customer Content transfer to idōs.

12.3 Capsule Memory and Sovereign Runtime Protections

All Capsule logic, Capsule Execution Memory, fallback structures, PRIS logs, AERA scores, sovereign metadata, and trust-governance artifacts are exclusive property of idōs.

Customer may view Capsule outputs or metadata through permitted tools but may not:

extract, replicate, or migrate Capsule logic;

retain Capsule Memory structures outside the Platform;

deploy Capsules or fallback vectors outside a licensed idōs runtime.

12.4 Customer Creations

Customer owns custom workflows, business logic, datasets, and Capsules authored by Customer (“Customer Creations”), provided they operate within the Platform.

However, the following remain the exclusive property of idōs:

Capsule scaffolds and fallback structures;

PRIS/AERA semantic infrastructure;

audit trails and sovereign-governance mechanisms;

runtime engines, memory architectures, and safety models.

Customer may not extract or reverse engineer these components or use Customer Creations to build external systems that replicate idōs functionality.

12.5 Aggregated Statistics

idōs may generate and own anonymized, aggregated metrics—such as Capsule invocation patterns, fallback frequencies, or entropy scoring trends—provided such metrics do not identify Customer or disclose Customer Confidential Information.

12.6 idōs Developments and Feedback

All inventions, enhancements, frameworks, algorithms, tools, simulation engines, or derivative works created by idōs (“idōs Developments”) remain the sole property of idōs.

All Feedback—including suggestions, critiques, improvement ideas, and runtime observations—are deemed assigned to idōs.
If assignment is not permissible, Customer grants idōs an irrevocable, perpetual, worldwide, royalty-free license to use and commercialize such Feedback.

12.7 Further Assurances

Customer agrees to execute documents or take reasonable actions to help idōs perfect its rights in Developments or Feedback. If Customer fails to do so, Customer appoints idōs as attorney-in-fact solely for that purpose.

12.8 Deliverables License

Subject to compliance with this Agreement, idōs grants Customer a limited, non-exclusive, non-transferable license to use Deliverables internally.
All underlying tools, Capsule scaffolds, algorithms, simulation modules, and sovereign-governance frameworks remain the exclusive property of idōs.

12.9 Reverse Engineering and Benchmarking Restrictions

Customer shall not:

reverse engineer, disassemble, decrypt, or derive source logic from Capsules, fallback chains, PRIS/AERA modules, Digital Twin engines, or agent behavior trees;

probe or train against outputs to deduce sovereign runtime logic;

use Platform outputs to build competing Capsule or DT systems;

benchmark, test, publish, or disclose runtime performance without prior written consent from idōs.

12.10 Confidential Information Overlap

All sovereign-runtime artifacts—including Capsule scaffolds, PRIS/AERA logic, fallback hierarchies, trust-governance metadata, and semantic outputs—constitute both:

Confidential Information (Section 11), and

Proprietary Rights (this Section),

and therefore receive dual protection.

12.11 Forensic and Audit Rights

Where idōs reasonably suspects IP misuse—including unauthorized replication of Capsule scaffolds or semantic governance structures—idōs may conduct a limited forensic review of Capsule access logs, trust-event histories, or runtime metadata strictly for verification and enforcement purposes.

12.12 AI Output and Derivative Work Boundaries

Customer may use semantic outputs and Results internally.
Customer may not:

assert ownership over sovereign runtime logic or Capsule scaffolding;

rehost, rebrand, resell, or redistribute Outputs externally;

use Outputs to train or tune external AI systems without written approval.

12.13 Frontier Model Firewall

Customer shall not train, fine-tune, mix, or inject idōs runtime data—including Capsules, PRIS/AERA logs, Drift metadata, or Digital Twin behavior—into external or open-source models, including but not limited to GPT, Claude, Gemini, or Llama, without written authorization.

This firewall ensures:

sovereign-runtime containment,

prevention of privacy or IP leakage, and

integrity of Capsule governance.

12.14 Enforcement Costs

If Customer breaches this Section, Customer shall reimburse idōs for all enforcement costs, including reasonable legal fees, expert analysis, forensic review costs, and remediation work.

12.15 Post-Termination Capsule Retention

Upon termination of the Agreement:

Customer may not retain Capsule scaffolds, runtime logs, semantic prompts, or fallback structures, except as expressly permitted;

idōs may revoke, quarantine, or regenerate Capsule environments to protect sovereign integrity.

12.16 Interoperability Boundaries

Customer may integrate the Platform with third-party systems provided that:

(a) all Capsule execution occurs only within licensed idōs runtimes;
(b) Outputs are not cached, recompiled, or transformed to mimic sovereign runtime behavior;
(c) cross-system Capsule calls remain governed under Platform constraints and are logged accordingly.

12.17 Twin-Generated Capsule Ownership

Capsules or behaviors generated autonomously by Personal or Enterprise Digital Twins are co-owned by idōs for governance and safety enforcement.
Customer may use such Capsules internally but may not publish, license, or export them outside the idōs ecosystem without written authorization.

12.18 No Compliance Extension to Third-Party Integrations

idōs provides governance, safety, and compliance assurances only within its sovereign runtime.
Any third-party systems, plugins, devices, or downstream applications connected by Customer are not covered by idōs’ warranties or compliance guarantees.
Customer is solely responsible for ensuring such integrations meet applicable laws and standards.

12.19 Survival

All restrictions, protections, ownership rights, and confidentiality requirements under this Section survive termination or expiration of this Agreement indefinitely.

Proprietary Rights – At a Glance

idōs Owns

Platform runtime, Capsules, PRIS/AERA logic, fallback trees, sovereign prompts, Capsule Repository

all underlying IP and idōs Developments

Feedback, aggregated statistics, and Twin-generated Capsules

all sovereign-runtime infrastructures

Customer Owns

Customer Content, Results, and Customer Creations

data and materials submitted to the Platform

internal workflows authored inside idōs

Shared / Licensed Rights

Deliverables (internal-use license only)

Outputs (internal use; no resale or external training)

You Cannot

reverse engineer or benchmark the Platform

export or rehost runtime artifacts

feed idōs outputs into external AI models

simulate or rebuild Capsule logic elsewhere

Enforcement

violations may trigger forensic review, quarantine, suspension, and recovery of full enforcement costs

these rights survive indefinitely

13. LIMITATION OF LIABILITY

13.1 Service-Level and Performance Limits

Customer acknowledges that Platform availability, Capsule responsiveness, semantic integrity, and agent containment are governed only by:

the service levels,

continuity protocols, and

fallback mechanisms

expressly described in this Agreement or in a signed Order / Support Agreement.
No additional uptime, performance, or quality guarantees apply unless set out in writing and signed by idōs.

13.2 No Consequential or Speculative Damages

To the fullest extent permitted by law, neither idōs nor its affiliates, licensors, upstream model providers, runtime partners, or Capsule contributors shall be liable, under any theory (contract, tort—including negligence—strict liability, product liability, or otherwise), for:

indirect, incidental, special, punitive, exemplary, or consequential damages;

damages for loss of data, semantic drift, incorrect or missed fallback actions, hallucinated Capsule outputs, or missed trust-score triggers;

damages arising from the use of the Platform in emergency, diagnostic, life-sustaining, legal, financial, or other high-stakes contexts;

loss of profits, revenue, business opportunity, goodwill, or business interruption;

costs of substitute goods or services,

even if idōs has been advised of the possibility of such damages.

13.3 Aggregate Liability Cap

Notwithstanding anything to the contrary, the total aggregate liability of idōs, its licensors, affiliates, and runtime providers, arising out of or related to:

this Agreement,

use or inability to use the Platform or Services, or

any Capsules, fallback maps, simulation trees, or embedded agents,

shall not exceed the total fees paid by Customer to idōs under the applicable Order during the six (6) months immediately preceding the event giving rise to the claim.

This limitation applies even if any remedy fails of its essential purpose.

13.4 Non-GA and Experimental Services

idōs has no liability for any outputs, errors, or damages arising from Customer’s use of Non-GA Services, including:

beta Capsules,

experimental agents,

prototype PRIS/AERA modules,

sandbox simulations.

Non-GA Services are provided strictly “as-is”, with no warranties, indemnities, service levels, or support obligations.

13.5 AI Evolution and Semantic Drift

Customer acknowledges that the Platform operates probabilistically and evolves over time through:

Capsule Repository updates,

safety patches and fallback recalibration,

drift-prevention routines, and

regulatory or governance compliance updates.

idōs shall not be liable for reliance on prior behavior, historical outputs, or legacy Capsule patterns that later change due to lawful or safety-driven updates.

13.6 No Liability for Data Loss or Corruption

idōs shall not be liable for loss, corruption, or delayed retrieval of:

Customer Content,

Capsule data,

Digital Twin logs, or

agent telemetry,

whether caused by outages, entropy anomalies, Customer misconfiguration, third-party hosting failures, or other external factors.
Customer is solely responsible for maintaining independent backups and redundancy outside the Platform.

13.7 Insurance Recommendation

Customer is strongly encouraged to maintain commercially reasonable insurance coverage—such as technology E&O, data protection, or cyber liability policies—to mitigate risks of AI usage.
idōs has no additional liability to the extent Customer fails to maintain such coverage.

13.8 Third-Party Dependencies and Frontier Models

idōs disclaims all liability for failures, errors, misbehavior, or defects arising from:

upstream third-party models,

open-source components,

external APIs, plugins, or integrations,

frontier LLM providers or hosting services.

Customer assumes all risk arising from reliance on such external services, including their retraining, deprecation, policy changes, or outages that affect Capsule outputs or DT behavior.

13.9 Cap on Remedies

Customer’s sole and exclusive remedies for any dissatisfaction with:

the Platform,

semantic outputs,

Capsules, agents, or Digital Twins, or

Support / Ancillary Services,

are those explicitly described in this Agreement and any applicable Order.
No additional implied, statutory, or equitable remedies apply.

13.10 Time Limit on Claims

Any claim, demand, or cause of action arising from or relating to this Agreement must be filed within twelve (12) months from the date the event giving rise to the claim occurred.
Claims not brought within this period are permanently barred.

13.11 No Liability for Autonomous Agent Actions

idōs shall have no liability for actions taken or not taken by:

autonomous Capsules,

Digital Twins, or

agents,

including self-modifying routines, fallback selections, or contextual reasoning sequences that operate within the sovereign runtime.
Customer is solely responsible for:

supervising such outputs,

validating them, and

determining their suitability before using them in production or human-facing contexts.

13.12 Enforcement of Risk Acknowledgment

Customer acknowledges and agrees that the limitations in this Section:

reflect a fair allocation of risk between the parties;

are an essential basis of the bargain for sovereign AI runtimes; and

apply even in cases of:

AI misfires,

incomplete or incorrect forecasts,

simulation errors, or

unexpected agent decisions.

13.13 Gross Negligence and Legal Carve-Out

Nothing in this Section shall exclude or limit liability for:

(a) gross negligence or willful misconduct;
(b) breach of specific contractual data-handling obligations expressly stated in this Agreement; or
(c) any liability that cannot lawfully be limited or excluded under applicable law.

Liability – At a Glance

idōs Liability Is Limited

Capped at fees paid in the last 6 months under the relevant Order.

No coverage for indirect, speculative, or consequential losses.

No liability for Non-GA features, data loss, third-party systems, or unsupervised agent actions.

You Remain Responsible To

Supervise and validate AI outputs.

Maintain backups and correct configuration.

Maintain your own insurance coverage.

Avoid using the Platform as a clinical device, emergency system, or fiduciary advisor.