Acceptable Use and First-Purchase Refund Policy

Revision: launch-aup-v4
Effective date: the date this revision is published on this page. It applies only after publication and Customer acceptance.

This Policy forms part of the Terms of Service and applies to the Customer, account owner, every named user and guest, every workspace, and anyone acting through Customer’s account. Capitalized terms have the meanings in the Terms. Customer is responsible for its Authorized Users.

1. Authorized use

Authorized Users may use supported clients interactively for internal Walmart operational reference and decision support within the account, workspace, seat, guest, and other limits in Customer’s plan or accepted Order. An advisor plan may be used to prepare work for a client represented in an authorized workspace. Authorized Users may access and use the Service only while physically located in the Launch Territory defined in the Terms of Service. Access is personal, nontransferable, and may not be rented, resold, or used as a service bureau.

Every paid plan includes every Retail Reason knowledge domain. A workspace’s channel and Scintilla tier affect what guidance applies; they do not authorize access to a third-party system or material.

2. Information you may submit

Submit only information that Customer lawfully controls and is authorized to provide to Retail Reason and the service providers identified in the Privacy Policy. Minimize all input to what is necessary for the question.

Do not submit:

• a Walmart, Retail Link, Supplier One, Scintilla, Marketplace, banking, email, or other password, session token, API credential, access key, or authentication code;
• a full payment-card number, card security code, bank credential, government identifier, health record, biometric identifier, precise location, child’s information, or other regulated sensitive data;
• unnecessary names, personal contact information, account numbers, transaction identifiers, raw sales files, or client-confidential data;
• Walmart-confidential or third-party material that Customer is not authorized to send to an AI service; or
• malicious instructions, code, files, links, or content designed to compromise, evade, or manipulate the Service or a provider.

Use a faithful summary, redacted excerpt, category, date range, or nonidentifying description whenever it will answer the question.

3. Prohibited conduct

You may not use or help another person use the Service to:

• violate law, sanctions, export controls, a court order, or another person’s rights;
• commit or facilitate fraud, deception, impersonation, harassment, discrimination, exploitation, threats, or unlawful surveillance;
• create, solicit, process, or distribute child sexual abuse material or other unlawful exploitative content;
• infringe or misappropriate intellectual property, privacy, publicity, confidentiality, contractual, or trade-secret rights;
• upload malware, probe or exploit a vulnerability, conduct penetration testing without prior written authorization, interfere with availability, overload the Service, or evade a security measure;
• access, enumerate, infer, or test another account, person, key, workspace, client, or provider resource;
• share a sign-in or access key, use another person’s seat, bypass a named-user or workspace limit, falsify an identity or billing record, or defeat revocation;
• bypass a rate, metering, review, refusal, evidence, or protective control, including through distributed accounts, users, sessions, prompts, or tools;
• scrape, crawl, batch, automate, systematically query, reverse engineer, decompile, or extract the Service, corpus, prompts, taxonomy, gated tables, source inventory, or protection thresholds except to the limited extent a restriction is prohibited by law;
• use the Service or Output to train, fine-tune, benchmark for replication, populate, or operate a competing model, dataset, knowledge base, advisory product, or answer service;
• resell, sublicense, republish, publicly distribute, or remove notices from the Service or a substantial collection of Output; or
• represent Output as official Walmart guidance, as independently verified professional advice, or as endorsed by Walmart, OpenAI, Anthropic, or another third party.

Do not use Output to make or materially influence a consequential decision about a person in employment, credit, housing, insurance, health care, education, legal services, government services, or another similarly high-impact area. Any exception requires Retail Reason’s prior written approval and every notice, assessment, human-review safeguard, and right required by applicable law.

4. Walmart and client boundaries

The Service grants no rights in Walmart systems, accounts, data, documentation, or marks. Users must comply with the Walmart, Sam’s Club, Retail Link, Supplier One, Scintilla, Marketplace, or other rules and confidentiality duties that govern their access. Retail Reason must not be used to scrape, systematically download, bypass, or automate unauthorized activity in those systems.

For every client workspace and guest, Customer must have documented authority from the relevant client and must limit access and information to that client’s work. A guest may use only the workspace granted to that guest. Attempting to discover another workspace, account directory, billing record, or aggregate usage is prohibited.

5. AI-assisted Output

Output may be incomplete, inaccurate, outdated, inconsistent, nonunique, or based on an unstated or incorrect assumption. A confidence, date, citation, evidence, or review label is not a warranty. Verify deadlines, monetary amounts, program rules, account facts, and operational steps in current official systems and governing agreements before acting. Use qualified legal, tax, accounting, employment, security, safety, or other professional advice where appropriate.

You may use and share Output as Sections 2 and 7 of the Terms permit, including sharing a complete delivered answer with the authorized client for whom it was prepared. You may not aggregate Output to reconstruct or replace the Service or corpus.

6. Usage status and extraction controls

The public account-level usage states are Normal, High usage, and Under review. High usage is an early warning. Under review means activity requires human review; it does not create an automatic fee, overage, or cost-based suspension. Retail Reason does not silently lower answer depth or quality because of account cost.

Separate person- and account-level controls may refuse a particular request and move activity to Under review when a pattern indicates systematic extraction, control evasion, credential abuse, or a security threat. We do not publish protection thresholds because doing so would make them easier to evade. A broader access restriction or termination follows Section 7.

7. Investigation, enforcement, notice, and review

We may investigate proportionately and preserve minimized evidence consistent with the Privacy Policy. Depending on severity, we may refuse a request, revoke or rotate a key, restrict a workspace, require corrective action, temporarily suspend an account, terminate access, preserve evidence under a legal hold, or report or cooperate where law requires.

Except where notice is prohibited or would materially impair security or an investigation, we will promptly notify the owner of the issue category, affected scope, effective time, available corrective action, and review channel. Emergency action may precede notice. For a curable issue, we ordinarily allow ten days to correct it. Deliberate extraction, credential compromise, fraud, unlawful conduct, cross-account access, or an immediate threat may justify action without a cure period.

The owner may request review within 30 days by emailing [email protected] with the subject “AUP Review.” Retail Reason will review the available evidence, correction, and proportionality and give a reasoned response. Because Retail Reason has one operator, this is not an independent appeal. Applicable statutory or card-issuer rights remain available.

8. Payment disputes and chargebacks

Contacting us about a suspected billing error is encouraged but is not a prerequisite to exercising a nonwaivable right. Customer may not seek duplicate recovery through both a refund and a payment dispute for the same amount.

We may temporarily restrict access supported by a disputed, unpaid, or credibly fraudulent charge while the dispute is pending. A pending restriction is not a finding that Customer acted improperly. Where lawful and safe, we will notify the owner, identify the affected subscription, and restore access if canonical payment evidence supports restoration. We will not penalize a customer solely for good-faith exercise of a nonwaivable card-issuer or statutory right.

9. One first-purchase refund per account

Each account receives one voluntary first-purchase refund opportunity. The eligible purchase is the account’s earliest successfully settled paid subscription purchase. It may be monthly or annual and includes the first paid conversion after an invite-only trial. A renewal, upgrade, plan-change increment, or later replacement purchase does not create another voluntary window.

Retail Reason must receive the owner’s request no later than 168 hours after the canonical settlement time displayed on the billing page. The deadline is inclusive. Receipt time, not the time we later review or process the request, controls. A request received one millisecond after the deadline is outside this voluntary policy, subject to any greater right required by law.

Submit through the authenticated billing page. If that channel is unavailable, send the request from the owner’s address to [email protected] with the subject “Refund Request” and include the account name and purchase date. Never send a password, access key, payment-card number, or other secret. We will preserve the original receipt time if a timely request requires later review or an approved retry after a provider failure.

The refund is for the full successfully settled payment attached to the eligible purchase. Access remains active while the provider reports the refund as pending, requiring action, failed, or canceled. When the provider confirms a full successful refund, paid access and keys end immediately and the subscription is canceled. Bank and card-network posting may take additional time and may appear as a reversal rather than a separate credit.

Suspected misuse may be investigated and sanctioned under Sections 6 and 7, but it does not erase the receipt time of a timely refund request. A refund may be withheld only where payment was not successfully settled, the request is outside this voluntary window, a refund has already succeeded, duplicate recovery would result, or law permits withholding because of proven fraud or abuse. We will give the owner a reasoned notice where lawful.

Nothing in this Policy limits a nonwaivable cancellation, cooling-off, refund, billing-error, warranty, chargeback, or other statutory right. A greater applicable right controls, and contacting us does not shorten a bank, card issuer, regulator, or court deadline.

10. Planned service closure

If Retail Reason plans to permanently discontinue the Service, it will provide the notice and unused-prepaid refund described in Section 15 of the Terms. That continuity refund is separate from the first-purchase refund and does not consume it.

11. Changes and contact

We may change this Policy only under the change process in the Terms. We preserve each accepted revision and its canonical text receipt. A material change will not retroactively convert previously permitted conduct into a breach.

Questions, refund requests when the authenticated channel is unavailable, notices, and review requests may be sent to [email protected]. Use a clear subject such as “Refund Request,” “AUP Review,” or “Security Report.”