This is a policy template for a merchant and its customers, not a sales contract between Aster and visitors to this website.
Store policies are templates for merchants to adapt and review against their actual seller, products and sales markets.
1. You are interacting with AI
Store example may use AI support for product, order and after-sales questions. The conversation is clearly labeled “AI assistant” at the start and does not impersonate an employee, lawyer or payment institution. Merchant example is responsible for its store’s support operation. AI is supplied through Aster and merchant-approved providers. This draft notice must match actual functions and data flows at launch.
2. Available help
AI may explain merchant-reviewed product, shipping and returns information, look up your order after necessary verification, organize an issue and draft an after-sales request. It accesses only information needed for the issue and must not show complete cards, other customers’ records or unnecessary recipient details. Without actual order or carrier evidence, it must say it cannot confirm rather than fabricate refund completion or delivery.
3. Human handling and limits
Every refund is approved by an authorized merchant employee. AI cannot alone transfer money, change a payout account or finally refuse legal requests. Withdrawal, disputes, product safety, privacy rights, payment anomalies and significant commitments must have human review. AI explanations do not change order prices, accepted terms or statutory protections. You do not need AI approval to submit a valid request.
4. Contacting a person and request timing
Choose Talk to a person in the conversation or contact support@example.com, telephone example. Human hours and timezone are example, and the usual response target is example. Repeated AI interaction, unnecessary explanations or an extra fee are not required. Requests can be submitted and recorded outside working hours, and legally effective notification is measured from valid submission rather than a later human handoff.
5. Data and sensitive information
Conversations may include your content, order references and handling records, used under the store privacy notice to answer, hand over and correct issues. Actual AI provider example, processing regions example and conversation retention example must be disclosed before launch. General-purpose model training is disabled by default. Do not send passwords, one-time codes, PANs, CVVs, private keys or unrelated sensitive information. Accidentally submitted secrets must be restricted and handled under procedure rather than added to model knowledge.
6. Errors and correction
AI may misunderstand or provide outdated information. You can refer a conversation to a person, and we check the actual contract, policy and order facts and explain corrections. The merchant cannot avoid responsibility merely because “AI said it.” Where legal interpretation is uncertain, AI must escalate rather than determine applicable law, consumer deadlines or liability limits itself.
7. Choice and continued access
Choosing not to use AI does not alter the substantive policies for buying, canceling, complaining or requesting refunds. Human channels must remain usable without unreasonable barriers. Disabling AI or a service outage does not erase submitted requests. Version: draft-2026-10-10. Operative date: example. Contact: support@example.com. Voice recording, screen sharing or new data purposes require separate disclosure and legally required choice or consent.