AI Features Terms

1. What these terms cover

The platform includes AI-assisted features — features that help generate or improve store content and product information, or that produce suggestions and recommendations for your store ("AI Features", and their results, "Output"). These terms set out what you can expect from AI Features, what we ask of you when you use them, and the limits of both.

As of 16 July 2026, and consistent with the feature descriptions published at /functionalities/, the AI Features generally available on the platform include:

• AI storefront search;
• generation of product descriptions, SEO titles and meta descriptions, and campaign copy;
• AI-assisted import of products, categories, and attributes from images, files, and tables;
• AI-assisted reports and recommendations, and content for Google, Meta, and TikTok;
• structured product data for search and AI-driven commerce.

Some capabilities are currently offered in beta or as partial functionality — at present the shopper and admin assistants and product AI-readiness scoring — and may change or be withdrawn. Features on our roadmap that are not yet marked available are excluded from this contractual inventory until they ship. The AI Features actually enabled for your account depend on your plan and configuration.

These AI Features Terms are incorporated by reference into the Master Subscription Agreement (Platform Terms) as a service-specific addendum. In case of conflict, your Order Form, the Master Subscription Agreement, and the Data Processing Addendum prevail, in that order; these AI Features Terms then prevail over general informational policies.

2. Probabilistic outputs — no guarantee of accuracy

AI Features are powered by machine-learning models that generate Output probabilistically. Output may be inaccurate, incomplete, outdated, biased, or misleading, and the same prompt may produce different results at different times. We do not warrant or guarantee that any Output is accurate, complete, reliable, current, non-infringing, or fit for any particular purpose. AI Features are provided "as is" and "as available"; nothing in these terms expands the warranties or liability provisions of your underlying agreement with us.

3. Your responsibility to review Output

• You are responsible for reviewing all Output before you publish it, rely on it, or use it to operate your store.
• Content you publish through the platform remains your content and your responsibility — including Output you choose to publish (product claims, pricing, availability, and legal statements such as ingredient, safety, or regulatory information).
• You must ensure that published Output complies with the laws that apply to you and your store — including consumer-protection, advertising, and product-safety rules — and with our Acceptable Use Policy.
• Do not present Output as authoritative fact without verifying it. Where Output includes factual claims about your products, verify them against your own records.

4. Prohibited high-risk uses

You may not use AI Features:

• To make, or as the sole basis for, decisions that produce legal or similarly significant effects on individuals — including decisions about employment, credit, lending, insurance, housing, education, healthcare, or access to essential goods, services, or legal services.
• To provide medical, legal, financial, or other professional advice to consumers, or in any safety-critical application (medical devices, emergency response, critical infrastructure).
• To identify natural persons (biometric identification), infer sensitive characteristics about individuals, or conduct surveillance.
• To generate content that is unlawful, deceptive, or infringing, or that violates our Acceptable Use Policy — including fake reviews, impersonation, and misleading product claims.
• To circumvent the platform’s safety measures or to extract, reverse-engineer, or benchmark the underlying models except as your agreement allows.

5. Our AI models — proprietary in-house AI agents

The AI Features are delivered by NEXT BASKET’s own proprietary AI agents — models and agent pipelines developed, operated, and trained in-house by NEXT BASKET. When you use an AI Feature, your prompts and the resulting Output are processed within our platform to deliver that feature; we do not route them to third-party foundation-model vendors as part of delivering the AI Features.

If we ever engage an external provider to help deliver an AI Feature, we will record its contracting entity, processing locations, data-retention settings, and its no-training / business-data terms, and list it in our subprocessor register at /subprocessors/ before it is enabled in the production environment. As of the effective date above, the AI Features run on our own in-house agents and no such provider is engaged.

6. Intellectual property in Output

As between you and us, and to the extent we hold any rights in Output generated for you, we assign them to you upon creation, subject to your agreement with us. However, the intellectual-property status of AI-generated content is unsettled: under current US Copyright Office guidance, purely AI-generated material is not copyrightable, and similar Output may be generated for other users. We therefore make no promise that you can obtain or enforce copyright or other exclusive rights in Output, or that Output does not resemble content generated for others.

You remain responsible for ensuring that your use of Output does not infringe third-party rights — including where Output incorporates or closely resembles existing works or trademarks. We do not provide an infringement indemnity for Output unless your written agreement with us expressly says otherwise.

7. Your data and model training

We do not use merchant content, shopper personal data, or website lead data to train general-purpose AI models unless the customer has expressly opted in through a separate written agreement.

This position also covers newsletter data collected through our website. Because the AI Features run on our own, in-house trained AI agents, your prompts and Output are processed within our platform to deliver the feature and are not shared with third-party model vendors to train their models. If we ever engaged an external model provider for a feature, we would first configure its API and business settings so it does not use your data to train its general-purpose models, and list it at /subprocessors/. Any opt-in to training would be a separate, explicit written agreement — it is never bundled into ordinary acceptance of the service, and never a default.

8. Personal data in prompts

• Do not input personal data of your shoppers or other individuals into AI Features except as strictly needed for the feature and as permitted by your Data Processing Addendum with us.
• Never input sensitive personal data (health, biometric, government IDs, precise geolocation, data about children) into prompts.
• You are the controller/business for personal data you submit; our processing of it is governed by the Data Processing Addendum. The AI Features run on our own in-house agents, so no third-party model provider receives your prompts; if an external provider were ever engaged, it would act as a subprocessor under the Data Processing Addendum and appear at /subprocessors/ before enablement.

9. Automated pricing, recommendations, and suggestions

Where AI Features suggest prices, promotions, product recommendations, inventory actions, or similar commercial decisions, these are suggestions to you, not decisions made for you. You decide whether to apply them, and you remain solely responsible for the prices, offers, and claims your store presents to shoppers — including compliance with pricing, anti-discrimination, and consumer-protection laws. We are not liable for commercial outcomes of applying Output.

10. Logging and retention

We may log AI Feature usage for security, abuse prevention, feature operation, debugging, service quality, and support. By default, AI operational logs consist of technical metadata and transient prompts and Output, retained for up to 30 days — unless you save content as part of your store, or a security or support incident requires us to keep it longer. Access to these logs is limited to authorised personnel, and because the AI Features run on our own in-house agents, these logs stay within our own infrastructure, where this retention position is enforced — no third-party model vendor holds them.

Logs are handled under our Privacy Policy and, for merchant personal data, the Data Processing Addendum.

11. No professional advice; AI disclosure

Output is not legal, medical, tax, financial, or other professional advice, and AI Features are not a substitute for a qualified professional. When you interact with an AI Feature — including any assistant or chat interface — you are interacting with artificial intelligence, not a human.

Every conversational or generative AI interface we ship clearly discloses, up front, that you are interacting with artificial intelligence and that its Output requires human review, and merchants who expose AI-powered interactions to their own shoppers should do the same. The current launch is B2B-only; we do not operate a separate U.S. consumer AI programme, and any U.S. consumer-facing AI-disclosure requirements, such as Utah’s, will be reviewed before a U.S. consumer launch.

You may not use any AI Feature to make consequential decisions about individuals — including in employment, credit, housing, insurance, healthcare, education, or access to legal services (see section 4). Readiness for Colorado’s and other high-risk-AI obligations is a 2027 compliance workstream; we document the required assessments before releasing any feature that could fall within their scope.

12. Changes and contact

We may update these terms as the AI Features and the law evolve; material changes will be notified per your agreement with us. Questions: office@nextbasket.com.