AI & Data Usage Policy

Last Updated: 11 January 2026

We respect the privacy of all users of the Eazli Platform and are committed to protecting personal data collected through www.eazli.com and its associated mobile applications (collectively, the "Eazli Platform").

This AI & Data Usage Policy explains how we use Artificial Intelligence ("AI") and data (including Personal Data) to provide and improve our services, including recommendations, personalization, Generative AI Agents, and our customer support chatbot.

This Policy should be read together with our Privacy Policy and Cookie Notice. Where there is a conflict, the Privacy Policy governs personal data matters.

What do we use AI for?

We may use AI systems (including machine learning models and automated tools) for the following purposes:

• Recommendations and personalization: suggesting products, content, or offers that may be relevant to you based on your activity and preferences.
• Search and platform optimization: improving search results, sorting, and the overall user experience.
• Generative AI Agents and customer support chatbot: using AI to respond to common questions, provide design and concierge assistance, and generate tailored guidance based on the information you provide.
• Analytics and service improvement: understanding usage trends and improving features and performance.

Generative AI Agents

Our Platform uses Generative AI Agents including, but not limited to (e.g., "Noura" for design and "Zeina" for concierge services). These Agents do not just answer questions; they can generate text, images, recommendations, and other outputs. These Agents may process your inputs (such as home photos, project descriptions, search queries, and chat messages) to create unique outputs and help you complete tasks on the Platform.

Because these Agents are automated, they may sometimes make mistakes, provide incomplete information, or generate outputs that are not appropriate for your specific situation. You should review and verify AI outputs before relying on them, and you can always request human assistance. Where an AI output could affect a transaction or another important decision, you retain final control and should apply appropriate human oversight.

Please do not share sensitive or highly confidential information via the AI Agents unless it is necessary for your request (for example, order number or shipping details).

You can request human assistance at any time by contacting our support channels.

Training vs. inference (how AI uses your data)

1. Inference: our AI features may process your inputs (e.g., search queries, browsing signals, home photos, project descriptions, and AI Agent messages) in real time to generate outputs, provide recommendations, and personalize your experience. This processing is used to deliver the service when you request it.
2. Improvement: we may use aggregated and/or de-identified data to evaluate, monitor, and improve the quality, reliability, and safety of our AI features (for example, reducing harmful or inaccurate outputs and detecting misuse).
3. User Content & Model Training (Privacy-First): we recognize the sensitivity of home-related information (for example, photos of your home interior, design files, blueprints, or private project details) and we use this information primarily to provide services to you. We do not use your private User Content to train public-facing AI models without your explicit consent. Where permitted, we may use anonymized, aggregated interaction data (including chatbot and AI Agent conversations) typically text-based — to improve the conversational abilities and overall performance of our AI Agents and to enhance quality and safety.
4. Third-Party AI Providers: we do not allow external AI providers to use your Personal Data to train or improve their models. If we use third-party AI services to help provide our features, they may process data only to deliver the service to us and must not use it for their own training, except as permitted by law and our agreements.

What data may be used for AI features?

Depending on your use of the Platform, AI-related processing may involve the following categories of data (to the minimum extent necessary):

• Account and profile data (e.g., username, preferences, preferred language).
• Purchase and transaction data (e.g., past orders, delivery/fulfillment information).
• Technical and usage data (e.g., IP address, device type, app version, logs, cookies and identifiers).
• Chatbot interaction data (e.g., chat messages, the time and context of the conversation, and any content you provide).
• Aggregated or anonymized data for analytics and improvement.

Sensitive data

In general, we aim to avoid processing Sensitive Data for AI purposes unless it is necessary for a specific service and permitted by applicable laws and regulations.

If Sensitive Data is required for a feature, we apply additional safeguards and obtain any required explicit consent.

Your choices (opt-out)

• You can disable personalization and recommendations through your account settings (where available). When disabled, you may still use the Platform, but some features may be less tailored to you.
• You can manage marketing preferences (where applicable) and cookie choices through the cookie banner and your device/browser settings.

Legal Basis

We process Personal Data for AI-related purposes only where we have a lawful basis under applicable laws, such as: performance of a contract, legal obligation, legitimate interests (where applicable), and consent (especially for direct marketing and certain uses that require consent).

Automated decision-making

Recommendations and personalization are generally used to assist your experience and do not produce legal or similarly significant effects by themselves.

If we ever deploy a decision that is based solely on automated processing and has a significant effect on you, we will provide required notices and safeguards, and obtain explicit consent where required by law, including offering a way to request human review where applicable.

Human-in-the-Loop: our AI Agents may assist with recommendations or draft responses, but they do not make binding decisions on your behalf. You always retain final control over transactions and commitments, and you can request human review or assistance where applicable.

AI governance & safeguards

We apply a risk-based approach to AI. Depending on the feature, we may implement controls such as:

• Human oversight for escalations, complaints, and sensitive cases.
• AI transparency: we clearly label our AI Agents so you can tell when you are interacting with an automated system versus a human support representative.
• Bias and fairness checks to reduce unfair outcomes.
• Quality and safety testing (including monitoring for harmful or incorrect outputs).
• AI limitations: While we strive for accuracy, AI outputs may be incomplete or incorrect and should not be relied upon as professional advice (for example, construction or code compliance) without verifying with a licensed professional.
• Explainability: where feasible, we may provide simple explanations for certain recommendations (e.g., "based on your preferences").
• Model change logs: we maintain internal records of material changes to AI models or configuration that may affect user experience.
• AI incident response: we investigate material AI-related incidents (e.g., unsafe outputs or systematic errors) and take corrective actions.

Sharing data with AI providers

We may share limited Personal Data with service providers (processors) that help us operate the Platform (e.g., hosting, analytics, payments, customer support tools, AI/chatbot infrastructure), subject to contractual obligations to protect your data and process it only on our instructions for specified purposes. We do not sell Personal Data.

Contact Us

For retention, security, breach notification, and international transfers, please refer to our Privacy Policy. This policy does not modify those terms.

If you have questions about this Policy or our AI features, please contact us at legal@eazli.com or the support channels listed on the Platform.

Changes to this Policy

We keep this Policy under regular review and may update it from time to time. Any changes become effective when we make the revised Policy available on or through our Platform.