AI Features and Acceptable Use Policy.

Scope and relationship to other terms

This AI Features and Acceptable Use Policy governs AI-assisted features in the Suntiq Fleet dashboard, Fleet mobile app, consumer mobile app, and related services. It forms part of the Website and App Terms. The Privacy Notice explains how Suntiq handles personal information.

A signed Customer agreement and data-processing addendum may govern Customer-controlled Fleet records. If those documents conflict with this Policy for a Customer's use, the signed Customer agreement controls to the extent of the conflict.

Covered AI features

This Policy covers:

• Ask and live voice conversations with Sofia;
• report drafting, narration, operational briefings, and scheduled organization narration;
• inspection-image comparison and proposed findings;
• market-valuation research and comparable-listing analysis;
• diagnostic narration and AI support chat; and
• other released features clearly identified as AI-assisted.

Deterministic calculations, stored Fleet records, and ordinary app controls are not converted into AI output merely because an AI feature may summarize them elsewhere.

Processing routes and providers

In the current service configuration, Ask, support, diagnostic, and report-narration requests go through OpenRouter to OpenAI. Inspection vision goes through OpenRouter to Anthropic. Valuation research goes through OpenRouter to Google. Live voice connects directly to OpenAI after Suntiq issues a short-lived session token.

When a text, narration, or inspection answer fails Suntiq's internal accuracy check, that request is retried once through OpenRouter on a Google model. The retry carries the same information as the first attempt, so Google is a disclosed provider for those features as well. The permission screen lists the current provider set before AI is enabled.

Every current OpenRouter request sets provider data collection to deny and disables silent provider fallback. Those controls restrict routing to providers OpenRouter classifies as not collecting request data; they are not an independent audit of a provider and do not establish blanket Zero Data Retention. A request can fail when a compliant route is unavailable.

OpenAI states that default API abuse-monitoring logs may contain customer content and may be retained for up to 30 days, subject to legal and safety exceptions. This applies to Realtime unless Suntiq later verifies an approved retention control for the exact account, project, endpoint, and model.

Suntiq does not currently run AI requests or output through a separate automated moderation or filtering service, and does not screen or verify every request or output. If Suntiq later enables a moderation provider, it will be named in the disclosed provider set and will carry the notice and renewed permission described below.

Information used and permission

Depending on the feature, a request may contain prompts, microphone audio, transcripts, photos you select, vessel identity and position, readings, alerts, maintenance, bookings, reports, prior conversation context, and other Fleet information needed to answer the request.

The mobile apps disclose the current provider set and require an affirmative action before sending personal information to third-party AI. Declining leaves core non-AI Fleet features available. A user can review or revoke that choice under Settings > AI data sharing.

A Customer administrator may decide whether the organization may use an AI feature under its Customer agreement and may control access to Customer records. An administrator cannot grant another person's mobile permission or authorize a person to submit information they do not have the right to disclose.

Some organization-directed processing is scheduled rather than started by a person — for example, narration attached to a completed rental or a periodic operational briefing. That processing runs on the Customer's authority, so an individual turning off AI data sharing in an app does not stop it. The in-app control governs requests that person's app makes.

The apps require renewed permission when the disclosed provider set or policy version changes. A provider change that does not affect a user's enabled features or disclosed processing may instead be handled through an updated provider register and notice where lawful.

Human control and source precedence

AI output is advisory. It may be inaccurate, incomplete, delayed, misleading, or unsuitable for the vessel, situation, or decision. Review the underlying Fleet records and other authoritative sources. Where AI prose conflicts with a deterministic calculation, stored Fleet record, confirmed human review, or qualified professional judgment, the non-AI source controls.

AI cannot independently:

• grant tenant, organization, location, vessel, or platform access;
• authorize safety-critical activity or emergency dispatch;
• control a vessel or operate autonomous navigation;
• make an insurance, underwriting, legal, or regulatory decision;
• confirm a diagnosis, repair, valuation, or inspection finding as fact; or
• execute an irreversible mutation or bypass a required human confirmation.

Prohibited use

You must not use AI features to:

• violate law, another person's rights, or a Customer's policies;
• create, request, distribute, or facilitate unlawful, exploitative, hateful, harassing, violent, or sexually abusive content;
• impersonate, defraud, deceive, stalk, dox, discriminate against, or endanger another person;
• obtain or expose confidential, personal, security, credential, or cross-tenant information without authorization;
• bypass authentication, rate limits, safety controls, human confirmation, or another product restriction;
• present AI output as verified telemetry, a confirmed inspection, professional advice, or a guaranteed outcome; or
• use AI as an emergency channel, sole navigation source, autonomous-control system, or sole basis for another safety-critical decision.

Reporting AI output

In the mobile apps, generated responses include an in-app Report action. A report always includes the displayed output and server-derived provenance. The prompt and relevant Fleet context are included only when the reporter selects the separate option. An optional note may also be included.

Suntiq records a durable report receipt and may use reports to investigate problems, adjust safeguards, and inform filtering or moderation. A report does not promise a particular outcome, response time, or automatic resolution.

Reporting is not an emergency channel. For an emergency, use VHF Channel 16 or the appropriate local emergency service.

Retention, deletion, and training

Suntiq may retain AI transcripts, conversation history, saved reports, consent events, and report events for service operation, Customer instructions, security, investigation, legal obligations, and claim needs. Device-local history remains until it is cleared, deleted, or removed through the app or operating-system lifecycle. Different records follow different schedules; Suntiq does not claim one universal retention period.

Account and data-deletion requests are handled as described on the Delete Account page and in the Privacy Notice. Deletion may not remove Customer-controlled records or evidence that must be retained under a contract, security need, legal obligation, or claim exception.

Suntiq does not use identifiable Customer-controlled content for unrelated model training without the required contractual authority, notice, and permission. Provider commercial terms and feedback paths may have separate rules described in the Privacy Notice.

Enforcement and changes

Suntiq may restrict or suspend AI features when reasonably necessary to investigate misuse, protect people or systems, comply with law, or enforce this Policy and the Terms. Severe or repeated misuse may lead to account or Customer action under the applicable agreement.

Suntiq may update this Policy prospectively. Material changes to disclosed AI processing, providers, user rights, or obligations will receive the notice and renewed permission required by law, platform policy, or contract. Prior versions are retained internally.

Contact

Suntiq, Inc 2093 Philadelphia Pike #3525 Claymont, DE 19703 Copy support email