AI Use Policy

Transparency

Customers should disclose AI-assisted interactions where required by law, by platform rules, or where a reasonable end user would expect to know they are interacting with AI or receiving AI-assisted support.

Customers are responsible for notices in their own channels, including websites, email, chat widgets, support portals, and customer contracts.

Human oversight

• Review AI-generated messages where accuracy, tone, compliance, safety, or customer impact matters.
• Require approval for domain transfers, DNS changes, billing changes, account changes, data deletion, bulk actions, or other sensitive workflows.
• Escalate low-confidence, unusual, legal, security, refund, complaint, cancellation, abuse, or high-value account matters to a qualified person.
• Monitor deployed workflows for accuracy, customer sentiment, escalation rates, false positives, and recurring edge cases.

Restricted uses

• Do not use Jexi for deception, impersonation, phishing, malware, spam, harassment, surveillance, unlawful discrimination, or rights violations.
• Do not use Jexi to make final decisions about credit, employment, housing, education, insurance, healthcare, legal rights, or similarly high-impact matters without qualified human review and written approval.
• Do not infer or target sensitive traits unless your agreement, notices, and laws expressly permit it.
• Do not bypass safety controls, approval workflows, rate limits, consent requirements, authentication, authorization, or audit logging.

Model training and customer content

Jexi does not use customer workspace content to train generalized AI or machine learning models.

Jexi does not use Google user data obtained from Google APIs to create, train, or improve generalized AI or machine learning models. Jexi also does not permit its AI model providers, subprocessors, or vendors to use Google user data to train or improve their generalized AI or machine learning models.

Customer-specific summaries, drafts, translations, workflow notes, and context records generated by Jexi are used only to provide the customer-authorized Jexi features inside the customer's workspace.

Accuracy and testing

Customers should provide accurate knowledge sources, test workflows before production use, maintain escalation paths, and periodically review output quality. Jexi output should not be treated as guaranteed fact without verification.

AI legal compliance

Customers are responsible for determining whether their use of Jexi is subject to AI, consumer protection, privacy, employment, accessibility, sector-specific, or platform-specific rules. Jexi may provide product controls and documentation, but customers remain responsible for their deployment decisions.

Reporting

Report unsafe, harmful, discriminatory, or unexpected AI behavior to hello@jexi.com.