ZexirAI · Terms of Service

The rules for using ZexirAI.

These Terms of Service form a binding agreement between you and the operator of ZexirAI. They explain who may use the service, what you may do with it, how subscriptions and user content are handled, and the limits that apply when software or AI output does not work as expected.

Plain-language note: AI output may be wrong, your account is your responsibility, and you must have the right to submit anything you upload. The detailed terms below control if this summary conflicts with them.

1. Agreement and scope

These Terms of Service (“Terms”) govern your access to ZexirAI websites, applications, chat interfaces, model-selection tools, voice and image features, file and project workspaces, subscription services, and related software, documentation, and support (collectively, the “Service”). “ZexirAI,” “we,” “us,” and “our” refer to the person or entity operating the Service. “You” means the individual using the Service and, when that individual uses the Service for an organization, the organization as well.

By creating an account, selecting a button that indicates acceptance, purchasing a subscription, or otherwise using the Service, you agree to these Terms and to the policies incorporated into them, including the Privacy Policy, AI Accuracy and Reliance Policy, Account and Security Policy, Subscription and Billing Policy, and Acceptable Use Policy. If you do not agree, do not use the Service. If you accept on behalf of an organization, you represent that you have authority to bind it.

Some features may display additional terms at the time you activate them. Those feature-specific terms apply in addition to these Terms. If there is a direct conflict, the more specific terms control for that feature, while these Terms continue to govern everything else. Mandatory rights under applicable consumer law are not waived by this agreement.

2. Eligibility and minors

You must be at least 13 years old to use ZexirAI, including guest features. Children under 13 are not permitted to use the Service, even with parental or guardian permission. ZexirAI has no upper age limit. If the law where you live requires a higher minimum age for a person to consent to an online service or to the processing of personal data, you must meet that higher age or use the Service only with authorization from a parent or legal guardian where the law permits it. Paid plans require you to be old enough to enter a binding payment agreement or to have the account holder’s permission.

ZexirAI is a general-audience productivity service and is not directed to children under 13. A parent or guardian who learns that a child under the applicable minimum age provided personal information should contact us so we can investigate and take appropriate action. Educational institutions and other organizations responsible for minors must independently determine whether use is appropriate, provide required notices, obtain necessary consents, and supervise use.

You may not use the Service if applicable law prohibits you from receiving it, if your account was previously terminated for serious or repeated violations, or if sanctions or export-control rules prevent us from providing access. You represent that registration details you provide are accurate and that you will keep them reasonably current.

3. What the Service does

ZexirAI provides an interface for working with artificial-intelligence models and related tools. Depending on availability, your plan, and your choices, the Service may route requests to models supplied by ZexirAI or by third-party providers, use web-search services, process files or images, generate images, transcribe or synthesize audio, preserve chat history, organize project files, export workspace data, or provide automated assistance for building and editing projects. Features, provider options, context limits, message limits, and supported file types may vary.

Model labels describe the experience offered through ZexirAI; they do not necessarily mean that the underlying model is owned or operated by ZexirAI. Provider names and trademarks belong to their owners. ZexirAI is an independent product unless a page expressly states otherwise. We may choose routing, fallback, safety, formatting, caching, or reliability systems needed to deliver a request. A selected model may be temporarily unavailable, replaced by an appropriate fallback, or limited by plan or capacity.

The Service is a tool, not a human professional, fiduciary, emergency service, public authority, medical device, law firm, financial adviser, or source of guaranteed facts. Details about appropriate reliance appear in the AI Accuracy and Reliance Policy. You remain responsible for deciding whether and how to use generated material.

4. Accounts and access

You are responsible for activity performed through your account, except to the extent it results from a security failure caused by us. Use a unique, strong password; safeguard authentication codes and recovery information; enable multi-factor authentication when available; and promptly tell us if you suspect unauthorized access. Do not share an individual account in a way that bypasses plan, security, or usage controls. You may not sell, rent, transfer, or lend an account without our written approval.

You must provide a working email address that you control. We may use it to verify registration, secure the account, send transaction or service notices, and respond to support requests. If you use Google or another sign-in provider, that provider’s authentication terms also apply. You authorize us to rely on valid credentials presented to the Service until you notify us of compromise and we have a reasonable opportunity to act.

Account controls may allow you to update profile details, export stored workspace information, delete individual chats or project files, manage multi-factor authentication, and view subscription status. Some requests, especially account deletion, ownership disputes, or changes to sensitive credentials, may require additional verification. More detail appears in the Account and Security Policy.

5. Your prompts, files, and projects

“User Content” means text, prompts, instructions, images, audio, files, code, feedback, project materials, and other information you submit to or create through the Service. As between you and ZexirAI, you retain your rights in User Content. You are responsible for it and represent that you have every permission, license, consent, and lawful basis needed for us and our service providers to process it as described in these Terms and the Privacy Policy.

You grant ZexirAI a worldwide, non-exclusive, royalty-free license to host, reproduce, transmit, format, analyze, and otherwise process User Content only as reasonably necessary to operate, secure, maintain, and improve the Service; fulfill your instructions; enforce these Terms; and comply with law. This license ends when the relevant content is deleted from active systems, subject to reasonable technical, security, backup, fraud-prevention, and legal-retention periods. It does not transfer ownership of your original content to us.

Do not submit secrets, regulated information, personal data, copyrighted works, confidential business materials, or another person’s information unless you understand the risks and are authorized to do so. “Private” mode prevents the conversation from being saved to your ZexirAI chat history, but the request still must be transmitted to our servers and applicable AI or search providers to produce an answer. It is not a promise of device-level secrecy, anonymity, or zero processing.

You may provide suggestions or product feedback. Unless you clearly mark material as confidential before we agree to receive it under a separate written obligation, you allow us to use feedback without payment or restriction to improve the Service. We will not publicly identify you as the source without permission.

6. Generated output and ownership

“Output” means text, code, images, summaries, suggestions, or other material generated by the Service in response to User Content. To the extent permitted by applicable law and subject to third-party rights, we assign to you any rights we may have in Output generated specifically for you. This does not grant rights in underlying models, software, interfaces, documentation, trademarks, or material owned by someone else.

AI systems may produce the same or similar output for multiple users. Output may incorporate common knowledge, generic patterns, public information, or material that resembles existing work. We do not guarantee that Output is unique, eligible for copyright or patent protection, non-infringing, accurate, or fit for commercial release. You must review Output, check sources and licenses, run security and quality tests, and obtain professional review when appropriate before publishing, deploying, selling, or relying on it.

You direct how Output is used and bear responsibility for that use. Do not represent AI-generated material as human-created where disclosure is required by law, contract, platform rules, academic standards, or professional ethics. Do not remove safety warnings or use Output to mislead others about evidence, identity, authorization, or expertise. The AI Accuracy and Reliance Policy gives practical requirements for high-stakes use.

7. Paid plans and billing

Free and paid plans may have different models, features, storage allowances, message limits, context limits, response speeds, and reset periods. Current plan details and prices are shown before purchase. Paid subscriptions renew automatically for the interval shown at checkout until canceled. By subscribing, you authorize our payment processor to charge the payment method associated with your billing account for the recurring price, applicable taxes, and charges you affirmatively approve.

You can schedule cancellation through the plan controls when available. Unless checkout states otherwise, cancellation takes effect at the end of the current paid billing period and access continues until then. Upgrading an active plan may take effect immediately and may produce a prorated charge or credit calculated by the payment processor. Downgrade paths may require canceling or using the hosted billing portal. Fees already paid are non-refundable except where we expressly agree or applicable law requires a refund.

We may change prices or plan contents prospectively. If a price change affects an existing recurring subscription, we will provide notice required by law before the change takes effect. Usage limits are protective service controls, not stored monetary value, and unused capacity does not roll over or convert to cash unless a plan expressly says so. The Subscription and Billing Policy is incorporated into these Terms.

8. Acceptable use and restrictions

You must follow the Acceptable Use Policy and all applicable laws. Among other things, you may not use the Service to harm or exploit people; create or distribute malware; steal credentials; facilitate fraud; evade sanctions; violate privacy, intellectual-property, or publicity rights; create non-consensual intimate content; sexualize minors; meaningfully enable weapons or violent wrongdoing; impersonate others deceptively; or make automated high-impact decisions without legally required review and safeguards.

You may not probe, scan, or test the Service for vulnerabilities without written authorization; bypass rate limits, plan gates, safety controls, authentication, or access restrictions; scrape or extract models, datasets, or large quantities of output through unauthorized automation; reverse engineer protected components except where law expressly allows it; interfere with other users; transmit malicious code; or use the Service to build a competing model through prohibited extraction.

Reasonable security research and ordinary interoperability work are not prohibited when performed lawfully, carefully, and within an authorized program or written permission. If you discover a vulnerability, do not access other users’ data, persist in systems, or publicly disclose details before we can investigate. Report it to support with enough information for safe reproduction.

9. Third-party services and links

The Service depends on third parties, including cloud infrastructure, authentication, payment, email, AI-model, search, and content-processing providers. Your request may be transmitted to a selected or routed provider to perform the function you requested. Those providers may impose their own terms and privacy practices. Our Privacy Policy describes categories of recipients and why information is shared with them.

Links, search results, plugins, generated code, integrations, and external content may take you outside ZexirAI. We do not control third-party sites or guarantee their availability, accuracy, security, legality, or continued compatibility. You are responsible for reviewing applicable terms before authorizing an integration, entering credentials, purchasing anything, downloading software, or allowing generated code to act on another service.

Third-party outages, policy changes, model deprecations, or restrictions may affect features. We may replace, suspend, or reconfigure a provider to preserve safety, quality, or availability. We are not responsible for a third party’s independent conduct, but we remain responsible for obligations the law places on us when we select and use service providers.

10. Service operation and changes

We work to keep ZexirAI useful and available, but the Service evolves. We may add, modify, limit, pause, or discontinue features; change models or providers; adjust storage or usage limits; release beta functions; correct errors; or perform maintenance. We may do so without advance notice when necessary for security, legal compliance, provider availability, or service integrity. When a material change adversely affects a paid subscription, we will provide notice or a remedy when required by law.

Beta, preview, experimental, and early-access features may be incomplete, change quickly, lose data, or be withdrawn. They are provided for evaluation and should not be used for production-critical or high-stakes work unless we expressly state they are ready for that use. You should maintain independent copies of important content and exports. The Service is not a backup system and no storage method is immune from failure.

We may set technical limits to protect capacity, manage costs, prevent abuse, or provide fair access. A displayed limit or estimate may lag briefly behind actual usage. Attempts to circumvent limits may result in throttling or suspension. We may communicate operational notices through the Service, email, or a status page.

11. Suspension and termination

You may stop using the Service at any time. Canceling a paid subscription stops future renewals but does not itself delete your account or stored content. You may delete individual content through available controls, export supported account data, or contact support to request account closure. We may need to verify identity and resolve an active subscription, legal hold, fraud concern, or ownership dispute before completing a request.

We may restrict, suspend, or terminate access when we reasonably believe you violated these Terms, created risk or harm, used the Service fraudulently, exposed us or others to legal liability, failed to pay amounts due, or threatened the integrity or availability of the Service. We may act immediately for urgent security, safety, or legal reasons. In other cases, we may give notice and an opportunity to correct the issue when practical.

Upon termination, your right to use the Service ends. Provisions that by their nature should survive—including ownership, payment obligations, disclaimers, liability limits, dispute terms, and rights needed to wind down or comply with law—remain effective. Content handling after closure is governed by the Privacy Policy and applicable law.

12. Disclaimers

To the maximum extent permitted by law, the Service and all Output are provided “as is” and “as available.” ZexirAI disclaims implied warranties of merchantability, fitness for a particular purpose, title, non-infringement, quiet enjoyment, and warranties arising from course of dealing or usage of trade. We do not warrant that the Service will be uninterrupted, error-free, secure, compatible with every device, or free of harmful components, or that content will never be lost.

We do not warrant the accuracy, completeness, originality, timeliness, safety, legality, or usefulness of Output, search results, citations, translations, code, images, or recommendations. AI can fabricate sources, misunderstand context, reproduce bias, omit material facts, or produce unsafe suggestions. You must independently evaluate all Output and use qualified human review for medical, legal, financial, employment, housing, education, insurance, safety, engineering, security, and other high-impact matters.

Some jurisdictions do not allow particular warranty exclusions. In that case, the exclusions apply only to the extent allowed, and you retain any non-waivable rights.

13. Limitation of liability

To the maximum extent permitted by law, ZexirAI and its suppliers, licensors, service providers, personnel, and affiliates will not be liable for indirect, incidental, special, consequential, exemplary, or punitive damages; lost profits, revenue, goodwill, opportunities, or data; business interruption; substitute services; or losses caused by reliance on Output, even if advised that such harm was possible.

To the maximum extent permitted by law, the combined liability of ZexirAI and the parties listed above for all claims relating to the Service or these Terms will not exceed the greater of one hundred U.S. dollars or the amount you paid to ZexirAI for the Service during the twelve months before the event giving rise to the claim. This cap applies collectively, not per incident, and regardless of the legal theory asserted.

These limits do not apply where prohibited by law, and they do not exclude liability that cannot lawfully be limited, such as liability for intentional misconduct or another category protected by mandatory law. The allocation of risk in this section is an essential basis of the agreement and applies even if a remedy fails of its essential purpose.

14. Disputes and general terms

Before filing a formal claim, please contact us and describe the problem, the account involved, the remedy requested, and relevant dates. Both sides agree to make a good-faith effort to resolve the dispute informally for at least thirty days, unless delay would cause a claim to expire or urgent relief is reasonably necessary. Nothing here prevents you from contacting a regulator or exercising a right that cannot be waived.

Applicable law governs these Terms, without displacing mandatory protections available in your place of residence. Any court with lawful jurisdiction may hear a dispute when the parties cannot resolve it. These Terms do not create a partnership, employment, fiduciary, franchise, or agency relationship. You may not assign them without our written consent; we may assign them as part of a merger, reorganization, financing, sale of assets, or transfer of the Service.

If a provision is unenforceable, it will be narrowed only as much as needed and the rest will remain effective. A failure to enforce a provision is not a waiver. Headings are for convenience. These Terms and incorporated policies are the entire agreement about the Service unless you and ZexirAI sign a separate written agreement. Notices may be delivered electronically.

15. Changes and contact

We may update these Terms to reflect new features, providers, business practices, security needs, or legal requirements. The effective date at the top shows when this version took effect. If a change materially affects your rights or obligations, we will provide notice appropriate to the change, such as an in-product message or email. Changes apply prospectively unless law requires otherwise. Continued use after the effective date means you accept the revised Terms.

Questions about these Terms may be sent to mail@zexiry.com. Include only the information needed to identify your account and explain the request; do not email passwords, authentication codes, full payment-card numbers, or highly sensitive content.

1. Account Eligibility: detailed guidance

This section explains how the policy applies to account eligibility. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on account eligibility, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to account eligibility. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns account eligibility, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of account eligibility as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

2. Service Access: detailed guidance

This section explains how the policy applies to service access. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on service access, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to service access. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns service access, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of service access as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

3. Guest Sessions: detailed guidance

This section explains how the policy applies to guest sessions. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on guest sessions, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to guest sessions. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns guest sessions, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of guest sessions as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

4. Paid Subscriptions: detailed guidance

This section explains how the policy applies to paid subscriptions. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on paid subscriptions, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to paid subscriptions. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns paid subscriptions, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of paid subscriptions as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

5. Model Routing: detailed guidance

This section explains how the policy applies to model routing. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on model routing, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to model routing. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns model routing, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of model routing as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

6. Third-Party Providers: detailed guidance

This section explains how the policy applies to third-party providers. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on third-party providers, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to third-party providers. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns third-party providers, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of third-party providers as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

7. User Content: detailed guidance

This section explains how the policy applies to user content. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on user content, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to user content. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns user content, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of user content as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

8. Generated Output: detailed guidance

This section explains how the policy applies to generated output. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on generated output, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to generated output. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns generated output, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of generated output as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

9. Search Results: detailed guidance

This section explains how the policy applies to search results. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on search results, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to search results. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns search results, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of search results as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

10. File Uploads: detailed guidance

This section explains how the policy applies to file uploads. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on file uploads, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to file uploads. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns file uploads, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of file uploads as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

11. Local Models: detailed guidance

This section explains how the policy applies to local models. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on local models, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to local models. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns local models, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of local models as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.

12. Plugins And Connectors: detailed guidance

This section explains how the policy applies to plugins and connectors. The governing principle is that a user should understand the feature, use only the access and information reasonably needed for the task, and keep qualified human judgment in control whenever an action may affect another person. Product labels, examples, interface text, and automated notices illustrate ordinary operation; they do not replace the specific facts, laws, contracts, professional duties, or provider terms that apply in a particular situation.

Before relying on plugins and connectors, review the requested purpose, the sensitivity of the information involved, the people who may be affected, and the consequences of an error. Use the least sensitive input that can complete the task. Confirm names, dates, amounts, permissions, sources, and destinations before sending, publishing, purchasing, deleting, or connecting anything. If a result appears incomplete, inconsistent, unusually confident, or outside the expected context, pause and verify it through an independent and authoritative source.

ZexirAI may apply technical limits, safety checks, rate controls, account verification, logging, provider routing, and human review to support this policy in relation to plugins and connectors. These measures reduce risk but cannot identify every mistake or misuse. A control that does not block an action is not approval, legal advice, a guarantee of accuracy, or a waiver. Users remain responsible for instructions they provide, content they choose to keep, actions they authorize, and the final use of any output.

When an exception, dispute, or unexpected result concerns plugins and connectors, preserve the minimum information needed to understand what happened: the approximate time, affected account, feature used, relevant settings, safe transaction or request identifiers, and a concise description of the expected and actual behavior. Remove passwords, one-time codes, full payment details, unrelated personal records, and confidential material. Contact support promptly when delay could increase harm, and use emergency or official reporting channels when the situation requires them.

We review the operation of plugins and connectors as the Service, models, providers, security threats, and legal requirements change. We may clarify instructions, adjust safeguards, restrict a capability, or request additional verification. Material policy changes receive notice when required. Existing non-waivable rights continue to apply. If this detailed guidance conflicts with a more specific section of this policy, the more specific section controls for that issue; the Terms and applicable law control the overall relationship.