AI Native

Privacy policy

1. Who we are and what this covers

AI Native is an organizational work platform operated by 1E6 Labs L.L.C-FZ, a company incorporated in Meydan Free Zone, Dubai, United Arab Emirates, at 6th Floor, Meydan Grandstand, Meydan Road, Dubai, United Arab Emirates (“we”, “us”). This policy covers the AI Native application at native.1e6.ai, its organizational assistants and enabled integrations. The service is for business use. The initial alpha is limited to admitted internal and friendly teams and ordinary work information. Do not supply regulated or highly sensitive information or data requiring a fixed deletion or in-country-only processing commitment that has not been expressly agreed and supported. The public website assistant and inquiry form have a separate website privacy notice.

For customer-controlled work data, the customer organization generally determines why the data is processed, and we process it to provide the service under its instructions and the applicable customer agreement. We determine how to process account, service-security and business-administration information for our own purposes where applicable law assigns us that responsibility. A customer’s own employee or privacy notice also applies to its workplace use.

2. Information we process

Account and membership information. When you sign in through Google or Microsoft, we receive the identity information needed to authenticate you, such as your provider identifier, name, email address and relevant organizational or tenant information. We also maintain your AI Native organization memberships, permissions and invitation state. Signing in does not itself authorize document or mailbox access.

Work information. We process information you and authorized colleagues enter or supply: conversations, requests, tasks, projects, organizational records, comments, uploaded files and generated results. This may include information about other people that your organization is authorized to use.

Connected-service information. If you separately connect a service, we process the permitted account identifiers, authorization grants, selected resources, resource metadata and content needed for the enabled functions. This can include document names, folders, provider references, revisions, permissions and document text. OAuth access and refresh tokens are credentials; they are held on the server and are not exposed as assistant responses.

Service and audit information. We process request times, security events, IP addresses and related connection information, errors, usage and credit consumption. Work actions and assistant activity also generate audit records, which can contain the relevant request or response content as well as action and permission metadata.

Support and commercial information. We process communications you send us and information needed to administer a customer relationship. Payment information, if a payment service is enabled, is handled according to the applicable checkout notice and customer agreement; this policy does not announce an active payment facility.

3. How we use information

We use information to authenticate users; maintain organization membership and access rules; organize work; generate answers and drafts; carry out authorized assistant actions; synchronize or retrieve admitted connected resources; operate background work that has been enabled; measure usage; support users; diagnose failures; protect the service; investigate abuse; and meet applicable legal obligations.

Assistants use relevant organizational context and enabled tools to perform these functions. Your instructions and permissions limit what an assistant can do. An organization’s administrator may enable or restrict capabilities, change permissions and choose available model providers. A permission to use an AI Native feature is separate from a provider’s consent to access its resources.

Where a legal basis is required, the applicable basis depends on the data and purpose: providing an agreed service, the customer’s lawful instructions, an applicable legitimate interest in operating and protecting the service, consent where required, or a legal obligation. An organization remains responsible for its own basis for submitting workplace and third-party personal information.

Information needed for authentication and organization membership is necessary to provide an account. Without it, we cannot provide that account. Optional connections are not required for basic sign-in; declining a connection means the functions needing that connection will not be available.

4. Google and Microsoft connections

Google documents. Enabled Google document connections use separate document consent and the Drive drive.file permission. This allows access to files created by the application or individually admitted to it, subject to the account’s own permissions. The document home is an approved folder in an organization-owned Shared Drive. The connection is not general access to every file in your Drive. Standard interactive access and agent access require separate clients and consent when both workflows are enabled. Google document connections remain unavailable until that separation has been implemented and tested; staged credentials alone do not enable access.

Microsoft documents. Enabled Microsoft document connections use delegated Sites.Selected permission and a separate grant for an approved SharePoint site. Access is constrained by both the site grant and the signed-in user’s permissions. Each customer uses its own Microsoft tenant and approved site.

Other integrations. Mailbox, calendar, social and other connections are optional and separately enabled. Their consent and in-product disclosures identify the data and actions requested. This policy does not enable an integration or expand a grant. In particular, signing in or connecting documents does not connect a mailbox.

You can disconnect a connection in AI Native and revoke the application’s grant in the provider’s account or administration settings. Disconnecting stops authorized future use of that connection; it does not automatically delete records, copied content, prior results or audit evidence already held in AI Native, and it does not undo an action already completed in the provider. The retention and rights sections below apply to those records.

5. AI processing and Google Limited Use

When an AI feature is used, relevant messages, organizational context and permitted source content can be sent to the model provider used for that operation. The shared alpha is limited to Anthropic, OpenAI and Google under their commercial API terms. DeepSeek and Moonshot are excluded from customer-data routes in this alpha. A fallback must stay within the approved providers. The enabled providers, processing purposes and location information are listed in the Data Processing Agreement. Infrastructure and model providers act under the applicable service and data-processing terms.

We do not sell Google API data, use it for advertising, or use or permit it to be used to train or improve generalized AI or machine-learning models. We use and transfer Google API data only to provide or improve the user-facing functions for which access was granted and for the other narrowly permitted purposes in Google’s policy. Human access is limited to the circumstances that policy permits, such as your affirmative agreement to access specified data, necessary security investigation, applicable legal obligations, or permitted aggregated internal operations.

Our use and transfer of Google API data follows the Google API Services User Data Policy, including its Limited Use requirements. Connecting Google does not authorize the use of its data for generalized model training. The same restriction applies to service providers receiving Google API data from us.

We do not use customer work content to train generalized models. The shared alpha uses paid Google API services and commercial OpenAI and Anthropic API services with provider training-sharing options off. Provider safety and abuse monitoring can retain requests under their applicable terms; this is not a zero-retention service. Model-provider selection and fallback must remain within these restrictions; choosing a model does not waive them.

Model-provider retention. Under their standard commercial API policies, OpenAI may retain abuse-monitoring logs containing prompts and responses for up to 30 days; Anthropic normally deletes API inputs and outputs within 30 days; and Google Gemini retains prompts, context and responses for 55 days for abuse monitoring. Legal, safety and policy-enforcement exceptions can extend retention. Optional stored conversations, uploaded files, caches, datasets or feedback can have separate retention periods. AI Native disables optional response or interaction storage where supported, but this does not remove the providers’ abuse-monitoring retention. We have not obtained a zero-data-retention arrangement for these accounts.

6. Who can receive information

Your organization. Authorized users and administrators can access information according to their organizational roles and the service’s access rules. Organizational records may remain available to authorized colleagues after the person who supplied them leaves. Organization administration does not itself grant unrestricted access to every personal connection or private conversation.

Service providers. We use infrastructure, storage, security, support and AI providers to operate the service. Microsoft Azure hosts the current shared estate. Google Workspace handles communications sent to our support address. Google and Microsoft process information for their respective sign-in services. Relevant data can also be sent to an enabled connected service when a permitted operation requires it. Providers receive information needed for their role, subject to applicable instructions and agreements.

Legal and business circumstances. We may disclose information when required by applicable law, to protect rights and service security, or in a business transaction subject to appropriate confidentiality and continuity safeguards. We assess requests and disclose only information justified by the applicable circumstances. We do not sell personal information or disclose it for targeted advertising.

7. Storage locations and international processing

The current shared application’s primary database and application storage are hosted in Microsoft Azure’s UAE North region, with geo-redundant recovery storage. Primary hosting is not the location of every copy or operation. Shared-tier AI requests can be processed outside the UAE. The approved model providers can process requests globally; the reviewed Anthropic workspace stores workspace data in the United States and uses global inference routing. Connected Google and Microsoft services and our support email also use their own service locations. Shared-tier use does not provide an in-country-only processing commitment.

Where applicable law requires safeguards for an international transfer, we use the arrangements required for the particular transfer, such as an applicable adequacy decision or appropriate contractual safeguards together with any required assessment. A request or connection consent does not replace a transfer mechanism required by law. Any customer-specific location or transfer commitment must appear in the customer agreement and match the enabled routes.

8. Retention and deletion

We retain information for the time needed for its purpose, the customer’s applicable instructions, security and evidentiary requirements, and applicable law. Different records have different lifecycles. Revoking a connection or leaving an organization does not erase its history.

The current default work-audit policy redacts request and response bodies after 30 days and removes audit rows after 400 days, through a scheduled sweep that may run up to a further day later. These windows concern that audit record class; they are not a promise to delete conversations, documents, organizational history, signatures or all copies of source content after 30 days.

The current infrastructure uses a seven-day database-backup window and 30-day storage soft-delete recovery windows. Storage versioning is also enabled. A current-file deletion is not an organization-erasure certificate. The file cleanup path must address older versions and failed cleanup attempts before all active bytes can be treated as erased. Technical logs have a separate default 30-day window.

Work history and derived context. Work records, conversations, copied source content and derived context are maintained to provide the customer’s ongoing workspace and history. They do not currently expire under the work-audit windows. Removal of a user or a connection does not remove the organization’s records. Customer-specific retention instructions and the arrangements on termination must be set out in the customer’s processing agreement. Requests for account, organization or retained-copy deletion are handled through the customer administrator and the contact below; an immediate self-service organization-erasure facility is not promised.

Connection credentials. Disconnecting a document account in AI Native deletes that user’s stored token bundles for the connection from the active database. Consent and disconnection evidence can remain, and credentials in recovery copies expire under the database-backup lifecycle. Disconnecting in AI Native does not itself revoke the provider-side grant, which you can separately revoke in Google or Microsoft.

Support, account and commercial records. We retain these for the time needed to resolve the request, administer the relationship, meet applicable accounting or legal obligations and handle relevant claims. We restrict retained information to those purposes. We do not treat these reasons as permission to retain every copy of customer work indefinitely.

Deletion and recovery copies. Where information must be erased, the process must address active records, derived copies and retained file versions as well as the current file. Recovery copies remain protected and unavailable for ordinary use until they expire or are removed. If a recovery copy is restored, applicable deletion instructions must be reapplied. Some information may instead be retained, restricted or de-identified where the customer’s lawful instructions or applicable law require it. We explain any applicable exception when responding to a request.

The proposed offboarding target is a 30-day export window and completion of active-data erasure or de-identification by day 60, with narrowly defined legal holds and the recovery-copy expiry windows above. This is an implementation target under review, not a currently supported contractual deadline. For this limited alpha, contact us to agree and carry out a return or deletion request; applicable statutory duties still apply.

9. Security and browser storage

We use organizational access controls, encrypted connections, protected server-side connection credentials and audit records. Organizations on the shared estate use shared infrastructure with organization-level separation enforced in the database. The shared tier does not promise a separate cloud tenant or dedicated infrastructure for each customer. No service can guarantee absolute security.

The application uses essential browser storage for sessions, sign-in state and interface preferences. Google and Microsoft sign-in interfaces may use their own cookies or browser storage. The current application does not implement advertising tracking. Any new optional tracking or materially different use will require an updated notice and any consent required by applicable law.

10. Your choices and rights

You can manage enabled personal connections and revoke provider consent. Your administrator manages organizational membership and permissions. Contact that administrator about customer-controlled workplace records, or contact us at support@1e6.ai for assistance, including access, correction, deletion, restriction, portability, objection or withdrawal requests where applicable law gives you those rights.

We may need to verify your identity and authority, and may refer a request concerning customer-controlled data to that customer. We respond within the period required by the applicable law and explain any lawful limitation. You may also complain to the competent data-protection authority. Withdrawal of consent does not invalidate earlier lawful processing.

Regional rights. Depending on where you are located and which laws apply, rights may arise under UAE personal-data protection law, Qatar's personal-data privacy protection law or applicable US state law. Business use does not itself remove individual privacy rights. You may use an authorized representative where the applicable law permits it. If that law provides a right to appeal a denied request, contact support@1e6.ai with the request reference and reason for appeal. We do not penalize you for exercising a protected privacy right. The service Terms' choice of governing law does not waive these rights.

11. Children and policy changes

AI Native is intended for business and professional use by adults aged 18 or older and is not available to children. If you believe a child has supplied personal information, contact us so we can investigate and take appropriate action.

We will publish the effective version and its date here. Material changes will be communicated through an appropriate service or customer channel, and additional consent will be obtained where required. Earlier versions and their effective dates will remain identifiable.

12. Contact

1E6 Labs L.L.C-FZ
6th Floor, Meydan Grandstand, Meydan Road, Dubai, United Arab Emirates
Privacy and data requests: support@1e6.ai
Customer support: support@1e6.ai

Version 2026-10-11 · Effective 11 October 2026