AI Usage Policy

This policy explains how AI-assisted features work on NexGen eHub and the responsibilities that come with them. It is part of our Terms of Service and applies alongside the Acceptable Use Policy and Privacy Policy.

The service is for people aged 18 or over and for business use. AI features are not designed for children, for personal use, or for clinical or medical purposes.

AI-assisted features include NexGen AI chat, AI Writer, product-description and listing tools, the business assistant, document reading and OCR, data preparation assistance, automation and workflow design, Voice AI and transcription, the Help Centre AI, and the Ask feature in Knowledge Studio. Some tools are deterministic — for example, profit calculators and compliance checklists — and do not use a generative model. The product page for each tool says what it does and what it does not do.

Features labelled limited, by enquiry or planned are described on the product pages. A tool being visible in the catalogue does not mean it is included in your plan; your plan and entitlements determine access.

Before an AI request is sent, the interface asks you to confirm that the relevant content may be processed. The confirmation is recorded with the request. Depending on the feature, the content sent can include:

• your prompt and the context you select;
• a sample of up to 20 rows from a CSV or spreadsheet, not the whole file;
• text or images from a document you ask us to read or extract;
• an audio recording you choose to transcribe or analyse;
• a question for the Help Centre AI and the matching published Help articles; or
• a web-research question and the search results returned for it.

We do not send your whole account, other customers’ content or files you have not selected. The Help Centre AI searches published Help Centre articles only; it does not search your Knowledge Studio articles or another customer’s workspace.

Our current AI provider is Google, using its Gemini API. The platform can be configured with other providers, and we may change or add a provider to keep the service working or improve quality. The providers in use for a contracted service are listed in the DPA schedule at the time you sign.

We do not use your content to train our own models. Providers process content under their own terms as our processor or sub-processor. Whether API content is retained or used for model improvement depends on the provider and the service tier. We are reviewing the current tier against our retention position; if you need the confirmed position before submitting particular content, contact us first.

AI requests use credits. The cost is shown before you run a request, and each charge is recorded in your usage history in the client portal. Plan allowances, renewal and top-up rules are set out in the Terms of Service.

• Plan credits renew at the start of each billing period; unused plan credits expire at renewal.
• Top-up credits do not expire while your account is active.
• Free-plan welcome credits are one-time and stay until used.
• If a request fails before a result is produced, we aim to release the reservation so you are not charged for an outcome you did not receive.
• Rate limits and fair-use controls protect the service. Automated bursts, bulk requests without a genuine business purpose, or attempts to use the platform as a general-purpose AI API are not permitted.

AI output can be inaccurate, incomplete, out of date or biased. It can invent facts, citations or product details, and it may not reflect your business, your customers or the law that applies to you. A confident tone is not evidence of accuracy.

We do not guarantee that an output is accurate, unbiased, secure, original or fit for a particular decision. We do not provide medical, clinical, legal, tax or financial advice through AI features, and no output should be treated as such. If a decision matters, verify it with an appropriate professional and your own records.

AI features are designed to assist a person, not to replace judgement. You should review every output before you send it, publish it, rely on it or use it in a customer interaction.

Do not use an AI output as the sole basis for a decision that has a legal or similarly significant effect on a person, such as a decision about employment, credit, housing, education, insurance or access to a service. Keep meaningful human review in the process. We do not make solely automated decisions about you, your access or your price.

Submit only the content needed for the task, and only content you are entitled to use. Do not submit special category data (such as health, biometric or political data), data about children, or confidential information you are not authorised to share. Our systems separate customer data, and AI features do not search another customer’s content.

Outputs and transcripts are saved only when you choose to save them. Account, billing and support data are handled as described in the Privacy Policy.

As between you and us, you may use the outputs you generate for your business, subject to third-party rights and the provider’s terms. We assign to you any rights we have in the output, but we cannot guarantee that an output is protectable by copyright, unique, or free of third-party claims. Similar outputs may be generated for other customers from similar prompts.

You are responsible for checking that an output does not reproduce a protected work or include another person’s personal data before you use it.

Where web research is enabled, the AI can search the web and return links to sources. Links may be wrong, paywalled, out of date or unrelated. Check the source itself before relying on a claim, and prefer primary sources for legal, regulatory or safety matters. The presence of a citation does not mean the answer is correct.

When you use Voice AI or transcription, the audio you select is sent to our AI provider for processing. We do not keep the recording after processing unless a feature explicitly saves it. A transcript is stored only when you choose to save it.

Transcription accuracy varies with accents, background noise, audio quality, overlapping speakers and technical vocabulary. Review the transcript before relying on it. If you record another person, you must have a lawful basis and, where required, their consent; do not use the feature for covert recording or to process another person’s audio unlawfully.

Data and document tools read the files you upload. Where an AI model is used, a sample of up to 20 rows may be sent rather than the whole dataset, and the interface tells you when a sample is used. Deterministic checks and calculations run on the data you provide and are not AI guesses.

Upload only files you are entitled to use, within the size and format limits shown. Do not upload malware or files designed to exploit a parser. Keep your own copy of important data.

You must not use AI features to:

• provide medical, clinical or diagnostic advice, or operate a medical device function;
• present AI output as professional legal, tax or financial advice;
• make a solely automated decision about a person with a legal or similarly significant effect;
• create deceptive content, fake identity documents, fake reviews or synthetic media presented as real;
• sexualise minors, promote self-harm, or give instructions for serious wrongdoing;
• attempt prompt injection, jailbreaks or extraction of another user’s data, system prompts or confidential information; or
• circumvent rate limits, credits or plan entitlements.

These restrictions sit alongside the Acceptable Use Policy. We may suspend access where we reasonably believe they have been breached.

AI features depend on third-party providers. They can be slow, unavailable or changed without notice, and a provider may impose its own limits. We may switch provider, model or region to keep the service working. We do not promise that a particular model, feature or output style will remain available, and we do not guarantee uninterrupted access. Where a provider outage affects a paid request, see section 5 on credits.

If an output is harmful, clearly wrong in a way that could cause loss, or appears to expose another customer’s data, contact us at info@nexgenehub.com with the subject “AI concern”, the feature used and, where possible, the request reference. Do not include sensitive content in the report unless it is needed to investigate. We will review the issue and tell you what we have done.

We update this policy when our AI features, providers or safeguards change. The “last updated” date shows the current version. Material changes will be announced in the client portal or by email before they take effect.