Terms of Service

These terms are a contract between Applied Physics Quantum Ltd. (“we”, “us”) and the organisation that signs up for QuinnAI (“you”, “your organisation”). By creating an account, joining an organisation, or using the platform, you accept them. If you are accepting on behalf of an organisation, you confirm you are authorised to bind it.

If we have signed a separate written agreement or order form with your organisation, that document governs where it conflicts with these terms.

Platform

The QuinnAI console, API, agent runtimes, integrations, and everything we make available as part of the service.

Organisation

The isolated workspace provisioned for your company, including its agent runtime, database schema, storage bucket, and vector namespace.

Customer Data

Everything you or your users put into the platform or connect to it, including prompts, conversations, knowledge documents, persona instructions, and records ingested from services you connect.

Output

What an agent produces or does in response to Customer Data, including generated text and actions taken through connected services.

Agent

An automated runtime that executes tasks on your behalf, inside the personas, permissions, and guardrails you configure.

The platform is in early access. Access is granted by invitation from the waitlist, and an invite code is bound to a single email address and can be used once.

During early access, features may change or be withdrawn, availability is not guaranteed, and some capabilities are labelled as coming soon. We will not deliberately destroy your data, but you should not treat the platform as your only copy of anything you cannot afford to lose.

You must give accurate registration details and keep your credentials secure. You are responsible for everything that happens under your account and under the accounts of your users.

Organisation roles are ranked: employee, manager, admin, owner. Owners and admins can invite and remove users, grant and revoke permissions, connect and disconnect integrations, configure personas and guardrails, and see organisation wide activity including task history, audit logs, and spend. If you are a user inside an organisation, understand that your work in the platform is visible to its administrators.

Every organisation must retain at least one owner. You are responsible for keeping owner access with someone who is still at your company.

Plans and pricing are agreed in an order form or invoice. There is no payment flow inside the product today, so nothing is charged to you automatically and no card is stored by us.

The platform meters model usage and estimated cost per task. Those figures are our best estimate for your visibility and planning. They are not a substitute for an invoice, and unpriced models are reported as unpriced rather than as free.

You can set daily and monthly spend caps. Caps are enforced on a best effort basis at the point a task is admitted. They can be exceeded by work already in flight, and they are not a guarantee of a maximum charge.

Fees are exclusive of taxes, which you are responsible for. Undisputed amounts are due as stated on the invoice.

You own your Customer Data. We claim no ownership of it.

You grant us a non exclusive, worldwide licence to host, copy, transmit, and process Customer Data solely to provide, secure, and support the platform for you, and to comply with law. That licence ends when the data is deleted.

We do not train, fine tune, or evaluate models on your Customer Data, and we do not use it to serve any other customer. Learning is opt in per interaction, and any memory it produces stays scoped to your organisation. The Privacy Policy describes this in detail.

You are responsible for having the rights and the lawful basis to put Customer Data into the platform, including personal data about your employees, prospects, and customers, and for telling those people as the law requires. Do not upload data you are not permitted to process.

As between you and us, you own the Output your agents produce, subject to the terms of the model provider that generated it. Because models are probabilistic, similar prompts can produce similar Output for other customers, and we make no claim that Output is unique or protectable.

Output can be wrong. Agents can misread a document, misuse a tool, or state something false with confidence. You are responsible for reviewing Output before you rely on it, and for any action taken on it.

The platform does not provide legal, financial, medical, tax, or other professional advice, and Output must not be treated as such. Do not use it as the sole basis for a decision with legal or similarly significant effects on a person.

Agents take real actions in real systems. They send messages, write to your CRM, and run on a schedule or a webhook without a human in the loop, if that is how you configure them.

The platform gives you controls: role and permission gating, approval steps, spend caps, per persona tool scoping, guardrail validation on agent output, and an audit trail. Configuring them appropriately for the risk you are taking on is your responsibility.

Those controls are safety measures, not warranties. Guardrail validation is pattern based and will not catch every case. You should require approval for any action whose consequences you would not accept if an agent got it wrong.

When you connect a service you authorise the platform to access and act on that account within the scopes you grant, on your behalf. Integrations can be connected once for the whole organisation or per user, and where the platform cannot resolve which person a turn belongs to, it refuses to spend a per user credential rather than guessing.

You are responsible for having the right to connect each account, for the actions taken through it, and for complying with the terms of that service. We are not a party to your relationship with a third party service, and we are not liable for its availability, its behaviour, or changes it makes.

Custom integrations you build with the platform’s tooling are yours to validate before you use them on live data.

The platform routes tasks to models you configure, which may include third party providers and models we host. Where a task runs on a third party model, the prompt and its assembled context are sent to that provider and that provider’s terms and acceptable use policy also apply to you.

Providers change models, deprecate them, and change their own terms. We may add, change, or remove available models and routing behaviour, and will give reasonable notice for changes that materially affect you.

You must not, and must not let your users or your agents:

• Break the law, infringe anyone’s rights, or violate a third party service’s terms.
• Upload data you have no right to process, or use the platform to process sensitive categories of data without an appropriate lawful basis and safeguards.
• Generate or distribute malware, spam, phishing, or content that harasses, defames, or sexualises minors.
• Use the platform to make consequential decisions about people, such as employment, credit, housing, or insurance, without meaningful human review.
• Attempt to reach another organisation’s data, probe or bypass isolation boundaries, or interfere with the platform’s security controls.
• Reverse engineer the platform, resell or sublicense access, or use it to build a competing service.
• Circumvent metering, spend caps, rate limits, or approval requirements.
• Present Output as human authored where doing so would deceive someone in a way that matters.

We do not routinely monitor your content. We may investigate when we have reason to believe these rules are being broken.

We own the platform, its software, design, documentation, and brand. Nothing here transfers those rights to you. You get a non exclusive, non transferable right to use the platform during your subscription, in line with these terms.

If you send us feedback or suggestions, we may use them without obligation to you. We will not identify you as the source without asking.

Each side may learn confidential information from the other. Each will protect the other’s confidential information with at least reasonable care, use it only for this agreement, and disclose it only to people who need it and are bound to keep it confidential. This does not cover information that is public, already known, independently developed, or lawfully received elsewhere. Disclosure required by law is permitted, with notice to the other side where allowed.

We may suspend an account, an organisation, or a specific capability if we reasonably believe it is being used in breach of these terms, if it threatens the security or stability of the platform, if it puts other customers at risk, or if payment is overdue. We will tell you and, where the cause is fixable, give you a chance to fix it, unless the urgency of the risk makes that unreasonable.

You may stop using the platform at any time. Either side may terminate for material breach that is not cured within 30 days of notice. We may terminate early access with reasonable notice.

Deleting an account in the platform console deactivates sign in and tears down the agent runtime behind every organisation that person owns. It stops access and stops spend. It does not by itself erase your records, your storage bucket, or your vector memory, which remain as the auditable history of what happened.

On termination, export what you need. We will delete or return Customer Data on written request to [email protected], subject to backups aging out on their normal cycle and to anything we must retain by law. Provisions that by their nature should survive, including ownership, confidentiality, disclaimers, liability, and indemnity, survive termination.

The platform is provided as is and as available. To the fullest extent the law allows, we disclaim all implied warranties, including merchantability, fitness for a particular purpose, non infringement, and any warranty arising from course of dealing or trade usage.

We do not warrant that the platform will be uninterrupted or error free, that Output will be accurate, complete, or suitable for your purpose, or that guardrails, sanitization, or spend caps will catch every case they are designed for. During early access we make no availability commitment at all unless one is written into your order form.

Neither side is liable for indirect, incidental, special, consequential, or punitive damages, or for lost profits, revenue, goodwill, or data, even if told such damages were possible.

Each side’s total liability arising out of or relating to this agreement is capped at the amounts you paid or owed us for the platform in the 12 months before the event giving rise to the claim. Where the platform is provided at no charge, that cap is one hundred units of the currency of your order form.

These limits do not apply to liability that cannot lawfully be limited, including death or personal injury caused by negligence, fraud, or fraudulent misrepresentation. They also do not limit your obligation to pay amounts due.

You will defend and indemnify us against third party claims arising from your Customer Data, your use of the platform in breach of these terms, actions your agents take through the accounts you connect, and your reliance on Output. We will tell you promptly of any such claim and let you control the defence, and you will not settle in a way that admits fault on our behalf without our consent.

We may change the platform and these terms. For material changes to these terms we will give notice in the console or by email before they take effect. If you keep using the platform after that, you accept the new version. If you do not accept it, stop using the platform and contact us.

These terms are governed by the laws of [jurisdiction], without regard to conflict of law rules. The courts of [jurisdiction] have exclusive jurisdiction, and both sides submit to them.

Before filing anything, please contact us. Most disputes are faster to resolve by talking.

These terms, plus any order form and the Privacy Policy, are the entire agreement between us on this subject. If a provision is unenforceable, the rest stands. Not enforcing a right is not a waiver of it. You may not assign this agreement without our consent, except to a successor of your business. We may assign it to an affiliate or a successor. Nothing here creates a partnership, agency, or employment relationship. Neither side is liable for delays caused by events outside its reasonable control.

Legal notices: [email protected]. Support: [email protected].

Applied Physics Quantum Ltd., [registered address].