Nativa Privacy Policy

1. Who we are

Nativa is a product family published and operated by NATIVA TECHNOLOGIES LIMITED, a New Zealand company (company number 9463088, NZBN 9429054000476). Nativa Technologies Limited is the controller responsible for the personal information described in this policy. Nativa Technologies Limited is wholly owned by its ultimate holding company, BECOMING LABS LIMITED (New Zealand company number 9436889), which is not the controller for Nativa.

2. What Nativa does

Nativa is a family of communication-assistance products across supported browser, desktop and mobile surfaces. It can help you translate messages, draft and refine replies, and understand messages, images, audio and documents that you explicitly ask it to work on. Features vary by product. Nativa's content-processing features run when you invoke them or when you enable a corresponding user-controlled setting.

3. Information we process

• Account and authentication data — such as your email address and session, account and membership state, where you use those features.
• Plan and billing administration data — your plan and subscription status, and limited Stripe identifiers and subscription metadata needed to administer a paid subscription for your account. Payments are processed by Stripe. Nativa does not receive or store your full payment card number or card security code; Stripe handles payment details under its own service and privacy terms.
• Product, device and service identifiers — identifiers assigned by Nativa to a browser installation, app installation, device session or account, together with structured membership and operational usage metadata needed to run, secure, meter and troubleshoot the service. This does not mean Nativa collects a hardware serial number.
• User-requested content — the message, selected or composed text you ask Nativa to act on and, when you explicitly invoke those features, images, audio and documents.
• Optional Communication Learning signals — structured behavioural signals, collected only while Communication Learning is switched on.

3A. Product analytics and quality signals

Operating Nativa includes content-free product analytics and quality signals. This is ON as part of providing, securing and improving the reliability of the service, and is separate from optional Communication Learning. It may include, at a bounded category level:

• random app-scoped installation, device and service identifiers;
• product, platform and app version;
• a coarse supported-app surface category (for example a messaging app type);
• the operation performed and its direction;
• the language pair;
• the feature or engine used (for example Fast or AI);
• whether conversation context was turned on (yes/no only);
• success or failure, with a fixed failure reason code;
• whether a retry or fallback occurred;
• latency and performance measurements;
• coarse input and output length buckets, never exact lengths;
• Replace, Send, Copy, regenerate and selection outcomes, where these can be observed reliably;
• plan and account state, only where determined by our servers.

This product analytics path does NOT store raw message, draft or clipboard text, translations, audio, transcripts, Conversation Context, contact or thread identifiers, the package names of host apps, hardware serial numbers, or precise location.

These identifiers are pseudonymous, not anonymous: they do not name you, but they can link events from the same installation, and where you are signed in our servers may associate them with your account. We treat them as personal information.

Product analytics does not switch on Communication Learning. Communication Learning remains explicit opt-in and OFF by default, and Aggregate improvement and raw content contribution remain separate higher-risk permissions that are never activated merely by product analytics.

4. Processing user-requested content

When you use a content-processing feature, including a feature you have configured to run automatically, the content needed for that feature is sent to the Nativa backend and to the AI provider selected for the requested feature and region so the result can be produced and returned to you. Google Gemini is used for supported regions generally. Alibaba Cloud Model Studio / Qwen is used for Mainland China, Hong Kong and Macau under current regional routing. Some translation paths may instead use translation available on your device or in your browser, or Google Translate. Content is processed to deliver the result you asked for and to operate and secure the service.

Nativa may use a coarse country or region signal from network or device context solely to select the appropriate AI provider. Precise GPS location is not required for this routing. Content routed to Qwen for Mainland China, Hong Kong or Macau is processed through Alibaba Cloud's China (Beijing) environment under the current configuration.

5. Communication Learning firewall

Communication Learning is explicit opt-in and is OFF by default. When it is on, the learning layer records only structured metadata — enumerated values, coarse buckets and version identifiers.

The learning layer does not persist raw message text, translations, drafts or replies, document contents, images, audio or transcripts, contact names, handles, phone numbers or email addresses, thread or conversation identifiers, or any free-form content you wrote.

6. Data Use Permissions

• Personalization is governed solely by your Communication Learning consent.
• Aggregate improvement is a separate explicit opt-in, OFF by default. If you explicitly enable it, it is limited to improving and evaluating Nativa's disclosed communication-assistance purpose, and nothing else.

No optional commercial or model-development permission is offered as a launch feature in the Chrome extension. Chrome extension user data — including derived, aggregated, anonymized or de-identified data — is never used for advertising, data brokerage, unrelated commercial datasets, or generalized model-development training outside Nativa's disclosed communication-assistance purpose.

7. Optional Content Contribution

Content Contribution is an optional explicit opt-in, OFF by default. It is separate from content-free product analytics, Communication Learning, Aggregate improvement, your Free, Plus or Pro membership, and ordinary use of Nativa. Text Contribution and Voice Contribution are independent scopes with independent controls. Enabling one does not enable the other, and contribution never applies retroactively.

For Text Contribution in Android V1, an eligible interaction is user-submitted outgoing AI Translate content using the AI engine without Conversation Context. It may include the source expression, Nativa AI output and final expression you edit or select, together with bounded provenance, language, tone and model metadata. Fast Translate raw content stays on your device and is not contributed.

Text Contribution excludes incoming, Read or Analyze source content; Conversation Context and conversation history; AI Reply training pairs; contacts, thread identifiers and host app package names; secure fields, passwords, PINs, one-time passwords and payment information; and background clipboard content.

Voice Contribution is a separate optional scope. It may include voice recordings and transcripts only when a supported Nativa product flow actually offers that contribution. Nativa does not currently provide a raw voice contribution path.

NATIVA TECHNOLOGIES LIMITED may use contributions to improve, train, fine-tune, distil, evaluate, benchmark, research, develop and commercialise current and future Nativa and related AI, language, speech, translation and communication technologies, models, datasets, derived datasets, products and services. This may create or improve corpora, model weights, embeddings and other derived technology. Contributions may be stored, copied, processed, curated, annotated and subject to privacy or de-identification processing.

Authorised personnel and contractors may review contributions. Our processors, cloud and model providers, and controlled research or commercial collaborators may process them only for the granted purposes. This processing may occur internationally in the countries where those recipients operate, subject to applicable law and our controls.

Contribution retention is different from the 180-day operational and 730-day quality and outcome periods for content-free product analytics. It is governed by the active contribution rights, the granted purposes and the applicable retention policy. Turning a scope OFF stops new collection for that scope. Withdrawing closes the contribution right and, after processing, prevents retained source contributions from new future dataset or training use.

Source deletion or account deletion removes or tombstones retained source contributions and closes related rights, subject to applicable law, records we must retain, and provider or infrastructure backups that age out on their own schedules. Withdrawal or deletion does not promise model unlearning or reversal of effects already lawfully incorporated into models, aggregates or derived technology.

If the Nativa business or Nativa Technologies Limited is acquired, reorganised, merged, financed or transferred, contribution data and rights records may transfer or remain with the Nativa business. A successor may continue use only within the purposes and rights originally granted and applicable law.

8. No sale of personal information

At launch Nativa does not sell personal information and does not deliver learning datasets to third-party commercial recipients.

9. Service providers and international processing

To operate Nativa we rely on service providers including cloud infrastructure, authentication and database infrastructure, and AI and translation providers. This means your information may be processed outside New Zealand, in the countries where those providers operate.

Our current providers include Google Gemini, for AI processing in supported regions generally and Google for some translation processing; Alibaba Cloud Model Studio / Qwen, for AI processing routed to Mainland China, Hong Kong and Macau; Supabase, for authentication and database infrastructure; Stripe, for payment and subscription processing; and Lovable, for application and backend hosting infrastructure. Raw user-requested content is not persisted in Nativa's learning layer, but that content can be transiently transmitted to and processed by Nativa's backend and these providers so the functionality you asked for can be performed. Provider and infrastructure systems may hold transient operational copies, logs or backups for their own limited operational periods.

10. Retention

Our currently configured retention for structured learning and operational data is:

• core behavioural signals — up to 730 days;
• derived behavioural signals — up to 730 days;
• operational metadata — scheduled retention of up to 180 days;
• operational product analytics — up to 180 days;
• content-free product quality and outcome signals — up to 730 days.

If you withdraw a learning authorization, the affected learning data is scheduled for removal under our current 30-day withdrawal retention policy. Raw user-requested content is not persisted in the learning layer. Infrastructure and provider systems may hold transient copies, logs or backups for their own limited operational periods, so this policy does not claim zero retention everywhere.

11. Your controls and rights

You can turn Communication Learning off at any time, and you can independently withdraw the higher-risk Data Use Permissions without turning off other features. Where in-product export and delete paths are available, you can use them directly in Nativa's settings.

You can delete your Nativa account in the product (for example in the Android app) or on the web at nativa.becominglabs.app/account/delete. Deletion removes your Nativa account and the Nativa data associated with it, and cancels any active subscription so it does not renew. Limited records may be kept where required for legal, tax, accounting, security or transaction purposes, and copies in provider or infrastructure backups age out on those providers' own schedules.

You can also request access, correction, deletion or export of your personal information by emailing info@becominglabs.app. Requests are handled subject to applicable law and reasonable verification of your identity.

12. Chrome Web Store Limited Use

Nativa's use of information received through the Chrome extension adheres to the Chrome Web Store User Data Policy, including the Limited Use requirements.

Allowed use. We use user data only to provide or improve Nativa's disclosed single purpose — helping you understand and express communications on supported websites by translating, explaining, and drafting or refining content in context — and for related operational purposes such as maintaining, securing, metering, measuring the reliability of, and troubleshooting those features.

Transfer. We transfer user data only when necessary to provide or improve that disclosed single purpose, to comply with applicable law, to protect against security threats, fraud or abuse, or as part of a merger, acquisition or sale of assets — and in that last case only after obtaining any prior user consent that is required. We do not transfer user data to advertising platforms or data brokers.

Human access. Humans are not allowed to read your content, except: where you give explicit consent for specific data to be read (for example, when you send it to us for support); where it is necessary for security or abuse investigation; where required to comply with applicable law; or for aggregated and anonymized internal operations consistent with applicable law.

Prohibited uses. We do not use user data for personalized, retargeted or interest-based advertising; we do not use it to determine creditworthiness or for lending purposes; and we do not sell it to data brokers or information resellers.

Raw user-requested content is not persisted in Nativa's learning layer. Content you ask Nativa to work on can still be transiently transmitted to and processed by Nativa's backend and its service providers solely to perform the functionality you requested.

Chrome extension data is not enrolled in Optional Content Contribution or used for generalized or commercial model-development merely because the same Nativa account enables a contribution scope in an eligible Android product. This carve-out also applies to Edge extension data, and the Chrome Limited Use commitments above are not widened by Content Contribution.

13. Security

We use HTTPS transport, access controls, and private backend and service-role controls that restrict privileged data paths. No online service can be guaranteed absolutely secure, and we do not claim otherwise.

14. Browser permissions

The Nativa browser extension needs access to the supported web services where it works so it can read the message you asked it to act on and place results back into the page. We do not use that access for advertising surveillance, ad targeting, or building profiles for advertisers.

15. Capacity and younger users

Nativa is a general-purpose product intended for people who have the legal capacity to agree to our Terms. If you are below the age at which you can agree on your own under the law that applies to you, you may use Nativa only with the permission and supervision of a parent or legal guardian.

16. Changes and contact

We may update this policy as Nativa changes. When we do, we will publish the updated version here with a new effective date and version identifier. Questions or privacy requests can be sent to the Privacy Officer, Nativa at info@becominglabs.app, or to our postal address above.

Nativa is available in all regions where it is distributed. Availability is not a claim of certification or compliance under every jurisdiction's law.

See also our Terms of Service.