AI and beta terms

These AI and Beta Terms supplement the Terms of Service between you and nith Studio LLC (nith). They apply when you use an AI-assisted feature, including Jot, writing or image tools, business setup or regeneration, brand or website analysis, and enabled automated screening. Sections concerning Beta Services apply only to features identified as beta, preview, experimental, or another pre-release designation when made available to you.

The Acceptable Use Policy applies to AI inputs, outputs, and actions. The Privacy Policy, Subprocessor Information, and any applicable Data Processing Addendum address processing of personal information. These Terms do not turn a planned feature into an available service or supersede non-waivable rights.

Input means prompts, instructions, references, files, and other material supplied or selected for an AI feature. Context means information retrieved or included by that feature to respond to the request, such as permitted site content, business settings, documents, selected records, conversation history, or enabled external sources. Output means generated text, images, designs, suggestions, analyses, classifications, and other results.

You are responsible for having the rights, authority, and legal basis needed to submit Input and authorize the Context used for your task. Do not submit credentials, prohibited sensitive data, or third-party confidential information you are not permitted to disclose. Information need not be published on your site to become Context: internal material that an enabled feature can access may be included in processing. Review the feature's scope and the privacy disclosures before enabling it.

By requesting a task or enabling an identified automated feature, you instruct nith to process the necessary Input and Context through the relevant service providers to provide that feature. This does not authorize unrelated disclosure or unrestricted reuse. Features available to your site may also include platform-configured automated checks, including screening incoming form submissions as described in the Privacy Policy. Such a check can process a submission when it arrives, without a separate prompt from you, and may not have a per-site off switch. You must provide notices and other protections required for the collection and processing supported by the features you use.

Different AI features have different effects. Some return advice or a proposal; some save drafts or generated assets; some apply an approved change; and some perform an enabled recurring or automatic task. Before starting or approving a task, review the feature's description, affected site or account, permissions, payer, and any confirmation it presents.

Generated setup pages and staged regeneration candidates require the publishing or replacement action indicated in those workflows. Other settings or operational changes may take effect when you confirm the relevant action, and generated material may be saved without being publicly published. Do not assume that every AI action is held in one universal draft or that everything affected by a build can be rolled back together.

Where a feature offers an approval step, you remain responsible for reviewing the proposed change before approving it. Where you enable an automated feature, you are responsible for the scope of that authorization and appropriate ongoing review. nith remains responsible for implementing its express feature commitments and applicable obligations; user review does not excuse a failure to honor those commitments.

A preview is a review tool, not a comprehensive security sandbox. External links, third-party integrations, and customer custom code can have effects outside a generated page preview. Version history, undo controls, and staged drafts operate only for the data and actions they expressly cover. Keep independent copies of important materials and verify a site's behavior before and after publishing.

AI systems can produce inaccurate, incomplete, outdated, biased, misleading, or inappropriate results, including convincing invented details, unsupported citations, incorrect business information, and problematic code or design. Similar results can be produced for different users. A fluent answer, a favorable score, or a successful generation is not verification.

Before publishing or relying on Output, use a qualified human reviewer appropriate to the subject and intended audience. Verify material facts, prices, product descriptions, contact details, sources, rights, accessibility, applicable disclosures, links, and functionality. Remove invented testimonials, credentials, claims, or endorsements. If a result concerns a regulated field, obtain appropriate professional review rather than relying on the AI feature as the professional adviser.

AI features are not emergency services and must not be relied on to direct immediate medical care or other safety-critical action. Do not use Output as the sole basis for decisions that have legal or similarly significant effects on a person, including eligibility for employment, credit, housing, insurance, education, or essential services. Any permitted use in such a context must independently satisfy applicable law, appropriate human oversight, and any expressly agreed service scope.

Automated lead screening and similar classifications may be wrong. Review affected records and maintain a suitable process for correcting errors. Scores from SEO, brand, or website analysis describe the checks or information available to that analysis; they do not guarantee rankings, legal compliance, measured browser performance, accessibility, conversion rates, or future commercial results.

You retain the rights you have in your Input. The limited processing license in the Terms of Service applies to Input, Context, and saved Output to the extent needed to provide the requested Services.

As between you and nith, and to the extent permitted by applicable law and our rights under the relevant provider agreements, nith assigns to you any rights it holds in Output generated specifically for you. This assignment does not include nith's or a third party's pre-existing software, templates, design components, stock assets, marks, or other independently licensed material that may be included in or used to produce Output. Those materials remain subject to their applicable licenses.

An assignment transfers only rights that exist and that nith can transfer. It does not establish that Output is protected by copyright, that you are its exclusive author, that it is unique, or that its use will not infringe another person's rights. Output generated for another user is not assigned to you merely because it resembles yours. Your rights against your own client or employer depend on your agreement with that party and applicable law, not merely on which account generated the material.

Before using Output commercially, evaluate intellectual property, privacy, publicity, and other permissions relevant to that use. Obtain releases or licenses where needed. Do not represent a generated image as a real event, a synthetic testimonial as a real customer's experience, or a generated professional claim as verified. Preserve attribution, license conditions, and provenance information that applicable law or an asset license requires. No intellectual property defense or indemnity is supplied unless expressly included in a signed agreement.

AI features use service providers identified in the applicable privacy and subprocessor disclosures. Depending on the feature, a provider can receive prompts, selected site or account Context, reference files, and technical information needed to process a request. Optional search, image, or other tools can involve additional providers or source websites as disclosed for that feature.

The provider's retention and permitted-use arrangements depend on the applicable service and configuration. We do not promise zero retention, a particular processing country, or a prohibition on every form of provider data use unless that commitment is expressly set out in the applicable agreement and supported for the feature. Information about our actual practices is provided in the Privacy Policy and applicable Data Processing Addendum. This clause does not authorize uses beyond those documents or displace mandatory data-protection requirements.

We may change models or service providers to maintain quality, availability, or safety, subject to applicable contractual notice, subprocessor, and data-protection requirements. Such changes can alter the style, speed, availability, and cost of future tasks. We do not guarantee access to a particular model unless an order expressly does so. Material changes to purchased functionality or fees remain subject to the Terms of Service.

The Terms of Service govern monthly and rolling credits. A feature may charge based on measured work or a fixed price shown before the task begins. Credits may cover processing even if you decide not to use or publish its result. Insufficient credit or an applicable task limit can prevent or interrupt a task. A safety refusal, provider outage, or failed tool step does not necessarily mean no processing occurred.

For a fixed-price feature, the description or confirmation identifies the charge and any automatic credit-restoration rule. For example, a regeneration that produces no pages and an eligible deep audit that fails may restore their charge, while completed work may remain chargeable even if discarded. The exact rule shown for the feature controls, subject to the refund policy and applicable law. Contact To be confirmed: ai billing support contact to report duplicate, incorrect, or otherwise disputed charges. This is not a blanket exclusion of refunds for defective service.

Where an agency or site can pay for a task, the initiating person must have authority to spend the selected account's credits. Review the displayed payer before starting. An agency's permission to edit a client's site does not automatically permit spending the client's funds or seeing its billing records. Site and agency balances are not a single shared balance unless the product expressly identifies them as such.

Beta Services are offered for evaluation and improvement while their behavior, limits, and interfaces may change. They may have defects, reduced support, incomplete integrations, or compatibility changes. Use them in a manner proportionate to the information involved and the consequences of failure, and test changes before relying on them for an important workflow.

A Beta Service is available only for the scope, period, and charges disclosed when you join or enable it. A beta label does not mean free access. Paid beta tasks, credits, and subscriptions remain subject to the disclosed purchase terms and refund rights. No launch date, continued beta access, or eventual inclusion in a paid plan is promised merely by making a beta available.

We may change or withdraw Beta Services with notice appropriate to the circumstances, subject to any express purchase commitment, applicable law, and the service-discontinuation provisions of the Terms of Service. Any special beta data-deletion risk, migration requirement, or reduced recovery capability must be disclosed before you elect to use the affected beta. This section does not authorize undisclosed deletion of customer production data.

The general confidentiality, personal-information, security, and liability provisions of the applicable agreement continue to apply. A beta designation does not create consent to train on customer data, waive a non-excludable remedy, or make all customer content public. If we ask for feedback, the feedback permission in the Terms of Service applies; participation does not oblige you to provide a public endorsement.

Do not circumvent model safeguards, platform access controls, or usage limits; submit data you are prohibited from disclosing; or use the AI features for conduct prohibited by the Acceptable Use Policy. Authorized research or testing requires the written permission described in that Policy.

You may stop requesting AI tasks or use an automated feature's off switch where one is available. Platform-configured checks do not necessarily have a per-site off switch. A control stops only the future tasks within its stated scope; it does not necessarily cancel work already in progress, remove saved Output, or delete provider-held records. Applicable cancellation, deletion, and retention rules continue to apply. Contact To be confirmed: ai support contact about available alternatives if a task or automated feature cannot be stopped through its interface.