AI Governance Policy

Cultara uses artificial intelligence, machine learning, classification systems, retrieval systems, and related analytical technologies to help customers understand organizational behavior, cultural patterns, and business-relevant risks. This AI Governance Policy explains how AI is used within Cultara and the safeguards we apply to support responsible, secure, and transparent use.

1. Scope

This Policy applies to AI-enabled features, models, workflows, analytics, classifications, dashboards, summaries, recommendations, and outputs provided through the Cultara platform.

It applies to Cultara’s use of AI in connection with customer-uploaded data, platform-generated outputs, and supporting operational processes.

2. AI Use Within Cultara

Cultara may use AI and related technologies to:

Classify anonymized or de-identified organizational communication patterns.
Identify cultural signals, behavioral patterns, and organizational dynamics.
Generate scores, summaries, dashboards, and analytical insights.
Support retrieval, ranking, and aggregation of relevant evidence.
Assist with trend analysis, risk identification, and executive interpretation.
Improve platform reliability, usability, and operational performance.

3. Responsible AI Principles

Cultara’s AI governance approach is based on the following principles:

Privacy-first design: AI workflows are designed to operate on anonymized, tokenized, or de-identified data wherever practicable.
Customer control: Customers determine what data is uploaded and how outputs are used within their organization.
Human oversight: AI outputs are intended to support human judgment, not replace it.
Transparency: We aim to explain the purpose and limitations of AI-enabled outputs in a clear and practical manner.
Security: AI processing is subject to platform security, access control, and tenant isolation measures.
Accountability: AI systems are governed through documented practices, internal review, and operational controls.

4. Data Used in AI Processing

Cultara is designed to process customer-prepared data. Customers are responsible for preparing, anonymizing, and uploading data in accordance with applicable laws and agreements.

AI processing may use:

Anonymized, tokenized, or de-identified communication-derived content.
Metadata provided or generated during preprocessing.
Functional group mappings or organizational categories supplied by the customer.
Assessment configurations, scoring frameworks, and customer-selected analysis parameters.
Platform-generated intermediate signals, classifications, summaries, and outputs.

5. No Training of Public AI Models

Cultara does not use customer data to train public AI models. Customer data is processed to provide the Service, generate customer-specific outputs, maintain the platform, and support authorized customer use.

Where Cultara uses third-party AI providers, Cultara configures such services, where controls are available, so that customer data is not used to train those providers’ public models.

6. AI Outputs and Limitations

Cultara outputs may include classifications, scores, summaries, evidence excerpts, dashboards, risk signals, trend indicators, and recommended areas for review.

Customers acknowledge that AI-generated outputs:

May contain inaccuracies, incomplete interpretations, or unexpected results.
Depend on the quality, scope, and representativeness of customer-provided data.
Should be reviewed in context by appropriate human stakeholders.
Are intended to support decision-making, not replace professional judgment.
Do not constitute legal, financial, employment, compliance, HR, or professional advice.

7. Human Oversight

Cultara is designed to support informed human review. Customers are responsible for determining how outputs are interpreted, validated, communicated, and used within their organization.

Customers should not rely solely on AI-generated outputs when making material decisions affecting individuals, employment, compensation, discipline, termination, legal rights, or other significant outcomes.

8. Individual Identification and Privacy Safeguards

Cultara’s platform is intended to identify organizational patterns and cultural dynamics, not to identify, profile, rank, monitor, or target specific individuals.

Customers must not use the Service to attempt to re-identify individuals from anonymized or tokenized data, infer individual identities, or make individual-level employment decisions based solely on platform outputs.

9. Bias, Fairness, and Context

AI systems may reflect limitations in source data, model design, configuration choices, analytical framing, or available context. Cultara seeks to reduce these risks through structured frameworks, defined scoring methods, review processes, and responsible design practices.

Customers are responsible for considering local context, organizational realities, legal obligations, and potential unintended effects when interpreting AI-enabled outputs.

10. Model and System Evaluation

Cultara may evaluate AI-enabled systems through internal testing, quality review, prompt and configuration review, output assessment, and monitoring of platform performance. These practices are intended to improve reliability, consistency, and usefulness of platform outputs.

Evaluation activities may use test data, synthetic data, anonymized examples, or customer-approved materials where appropriate.

11. Security and Access Controls

AI processing is subject to Cultara’s security controls, which may include tenant isolation, encryption, role-based access controls, monitoring, logging, and infrastructure safeguards.

Access to customer data and AI processing systems is limited to authorized personnel and systems with a business need to provide, maintain, secure, or support the Service.

12. Third-Party AI Providers

Cultara may use trusted third-party AI providers or infrastructure providers to support platform functionality. Such providers are used only as necessary to operate, maintain, or deliver the Service.

Where third-party providers process customer data, Cultara applies contractual, technical, and organizational safeguards appropriate to the provider’s role and the nature of processing.

13. Customer Responsibilities

Customers are responsible for:

Ensuring data uploaded to the Service is lawful, appropriate, and properly prepared.
Maintaining authority to process data, including when acting on behalf of clients or third parties.
Reviewing AI outputs before relying on them for material decisions.
Using outputs in a manner consistent with applicable laws, agreements, and internal policies.
Avoiding attempts to re-identify individuals from anonymized or tokenized content.
Ensuring that platform outputs are communicated responsibly and with appropriate context.

14. Prohibited Uses

Customers may not use Cultara AI-enabled outputs to:

Identify or attempt to identify specific individuals from anonymized data.
Make solely automated employment, disciplinary, compensation, termination, legal, or similarly significant decisions about individuals.
Conduct individual surveillance, ranking, profiling, or behavioral monitoring.
Violate applicable privacy, employment, human rights, anti-discrimination, or data protection laws.
Misrepresent AI-generated outputs as definitive facts without appropriate review and context.

15. Transparency to Customers

Cultara aims to provide customers with practical information about how AI-enabled outputs are generated, including the general nature of inputs, analytical frameworks, scoring methods, confidence indicators, evidence summaries, and limitations where applicable.

Some technical details, model configurations, security controls, or proprietary methodologies may remain confidential to protect platform integrity, security, and intellectual property.

16. Updates to This Policy

Cultara may update this AI Governance Policy from time to time to reflect changes in our platform, AI practices, legal requirements, or governance approach. Updated versions will be posted with a revised "Last Updated" date.

17. Contact

Questions regarding this AI Governance Policy may be directed to:

Email: connect@cultara.ai