Privacy Policy & Data Security
Your data is your business. We process it transparently, securely, and with clear controls.
Last Updated: August 3, 2026
1. Scope, Roles, and Applicability
This policy explains how Tuba Intelligence Limited processes personal data when you visit, register for, or use Tubagen. Depending on context, we may act as a data controller (for account and billing operations) or a data processor (when handling customer project data on behalf of a workspace owner).
2. Categories of Data We Process
Account and Workspace Data
Name, email, role, workspace membership, authentication records, support history, and account settings.
Social Login Data
If you register or sign in using a supported third-party login method, we may receive profile information needed to authenticate your account, such as name, email address, profile image, account identifier, and verification status.
Content and Project Data
Uploaded media, scripts, briefs, scene and shot structures, workflow nodes, prompts, parameters, generated outputs, saved assets, templates, tags, task records, review comments, and project metadata.
Technical and Usage Data
Device identifiers, browser and OS information, log files, session events, IP address, and basic analytics telemetry.
Commercial Data
Plan, billing events, invoices, payment status, and credits usage (handled by authorized billing providers and merchant-of-record services).
3. Why We Process Data (Legal Bases)
Contract Performance
To provide the service, authenticate users, generate outputs, maintain projects, and enable workspace collaboration.
Legitimate Interests
To secure the platform, prevent abuse, improve performance, and maintain product reliability.
Aggregated and De-identified Data
We may use aggregated, de-identified, or statistical usage data to improve service performance, reliability, analytics, and product planning. We do not treat aggregated data that cannot reasonably identify a person or workspace as personal data.
Consent
Where required, for specific processing activities such as optional marketing communications or model-training opt-ins.
Legal Obligations
To comply with applicable laws, lawful requests, accounting obligations, and dispute handling requirements.
4. AI Processing and Training Policy
AI Processing Infrastructure
When you run workflow nodes such as image generation, video generation, image or video editing, speech, music, sound effects, AI assistant, frame extraction, composition, or related AI tools, relevant data may be processed through approved AI infrastructure as required to deliver requested outputs.
Creative Agent on Amazon Bedrock (United States)
When the Creative Agent deployment is explicitly configured to use Amazon Bedrock, Tubagen sends the user's message text, the Creative Agent system instructions, and relevant prior messages from the Canvas-owned shared Thread—including messages written by other authorized collaborators—to Amazon Bedrock in the us-west-2 United States region. In Phase 1, image attachments are not sent to Bedrock. Bedrock and Ark are separate deployment profiles; requests are not automatically rerouted or failed over between them.
Bedrock Storage Setting and Provider Handling
Bedrock requests are sent with store=false. This setting is not a zero-data-retention commitment and does not establish that AWS retains no data, creates no logs, performs no safety or compliance review, or deletes data immediately. AWS may process or retain data under its applicable service terms, policies, and legal obligations.
Creative Agent on Volcengine Ark (Mainland China)
For the Creative Agent service made available within our mainland China scope, Tubagen may send the user's message text, the Creative Agent system instructions, and relevant prior messages from the Canvas-owned shared Thread—including messages written by other authorized collaborators—to Volcengine Ark at ark.cn-beijing.volces.com in Beijing, China. In Phase 1, image attachments are not sent to Ark. This processing occurs only when that deployment is explicitly configured to use Ark; Ark is not an automatic fallback for other regions.
No Training by Default
Tubagen does not use private uploads or customer project data to train AI models. We do not authorize our AI inference providers to use that data for model training unless we first provide an updated disclosure and obtain any consent required by law. For the currently configured Volcengine Ark account, both the data authorization agreement and collaboration reward plan are inactive. This training restriction does not prevent providers from processing or reviewing inputs and outputs to deliver the service or for logging, security, legal compliance, abuse prevention, and risk filtering as permitted by their terms and applicable law.
Ark Storage Setting and Provider Retention
Ark requests are sent with store=false. This is not a zero-data-retention commitment. The exact retention period, logging categories, and account-specific retention or logging boundary are currently unknown. Volcengine may retain, log, process, or review inputs and outputs for service operation, security, legal compliance, abuse prevention, and risk filtering under its service agreement, data protection terms, privacy policy, and applicable law.
Biometric and Likeness-Sensitive Data
If you upload real-person materials, authorization recordings, likeness, face, voice, or other sensitive identity material, you confirm you have the required rights and consents. You are responsible for lawfully obtaining permissions from any person whose likeness or voice is used.
5. Data Residency, Transfers, and Subprocessors
Infrastructure and Hosting
Depending on the explicitly selected Creative Agent deployment profile, AI inference uses Amazon Bedrock in the us-west-2 United States region or Volcengine Ark at its Beijing, China endpoint. A request is sent only to the Provider selected for that deployment.
International Transfers
The Creative Agent data described in Section 4 is processed by Amazon Bedrock in the us-west-2 United States region when the Bedrock deployment is selected, or at Volcengine Ark's Beijing, China endpoint when the mainland China deployment is selected. Where data is transferred across borders, we apply appropriate safeguards such as contractual clauses and equivalent protections required by applicable law.
Subprocessors
We engage vetted subprocessors for infrastructure, analytics, payments, communications, and AI inference. Amazon Bedrock acts as the AI inference provider for the Bedrock Creative Agent deployment, and Volcengine Ark acts as the AI inference provider for the mainland China Creative Agent deployment. We require contractual security and privacy commitments from each subprocessor.
Business Transfers and Corporate Changes
If we are involved in a merger, acquisition, financing, restructuring, or sale of assets, relevant data may be transferred as part of that transaction subject to appropriate confidentiality and legal safeguards.
6. Security Controls
Encryption and Transport Security
Data is protected in transit using TLS and protected at rest using industry-standard encryption controls.
Access Management
We apply least-privilege access controls, role-based permissions, and monitored administrative access to production environments.
Operational Security
We maintain logging, monitoring, incident response processes, and periodic control reviews to protect service integrity.
No Perfect Security
No internet transmission, storage system, or cloud service can be guaranteed to be completely secure. You should use strong credentials, protect account access, and access the service from secure environments.
7. Retention and Deletion
We retain personal data for as long as needed to provide services, meet legal obligations, and resolve disputes. Workspace owners can delete projects and assets, and may request account deletion subject to legal and security retention requirements.
8. Your Rights and Choices
Data Rights
Depending on your jurisdiction, you may have rights to access, correct, delete, restrict, object, or request portability of your personal data.
Hong Kong PDPO Rights
Where the Hong Kong Personal Data (Privacy) Ordinance applies, you may request access to personal data we hold about you, request correction of inaccurate data, and withdraw consent where processing is based on consent. We may take reasonable steps to verify your identity and may charge a lawful, reasonable fee for data access requests where permitted.
Workspace Context
If your data is processed under a customer workspace, requests may be routed to that workspace administrator as the primary controller of project data.
Marketing and Preferences
You can opt out of marketing emails at any time via unsubscribe controls or by contacting us.
9. Cookies and Similar Technologies
We use cookies and similar technologies to support authentication, security, session continuity, analytics, and product performance. You can control cookie preferences through browser settings where available.
10. Children and Age Requirements
Tubagen is not intended for children under 18 or under the age required by applicable law to consent to data processing, whichever is higher. If you believe a child has provided personal data in violation of this policy, contact us for removal review.
11. Policy Updates
We may update this policy to reflect legal, technical, or product changes. Material updates will be posted on this page with a revised effective date.
12. Contact Us
For any privacy-related inquiries, please reach out to:
Legal Department
Tuba Intelligence Limited
Registered Office:
Room 701, Unit 108, 7/F, Tower B
New Mandarin Plaza
14 Science Museum Road
Tsim Sha Tsui, Kowloon
HONG KONG
The address above is our registered office address for legal correspondence.
legal@tubagen.ai