Mosaic Terms of Service
Last updated: 5 October 2026
These terms are an agreement between Campainless Pty Ltd (ACN 668 774 856, ABN 18 668 774 856) of 83 Princes Highway, St Peters NSW 2044, Australia ("Campainless", "we") and the organisation that creates a Mosaic workspace ("you", "your organisation"). They govern use of Mosaic — the organisational-knowledge and AI-coworker platform at app.mosaicailab.ai, its Slack application, and its connections to your other tools ("the Service").
By creating a workspace, accepting an invitation into one as its administrator, or continuing to use the Service after a notified change to these terms, you agree to them on your organisation's behalf and confirm you are authorised to do so. These terms incorporate our [Privacy Policy] and [Acceptable Use Policy].
1. The Service
Mosaic builds a governed knowledge base ("the brain") from the sources your organisation connects, and provides an AI coworker that works over that knowledge and those sources — answering questions, running skills, executing scheduled work, and proposing changes to the knowledge base that your people approve or reject.
We will provide the Service with reasonable skill and care. We do not promise it will be uninterrupted or error-free, and at this stage of the product we do not offer a service-level agreement; where an SLA is agreed with you separately in writing, that document governs availability.
2. Your account and your people
- You are responsible for the people you admit to your workspace, for the accuracy of your account information, and for keeping credentials confidential.
- Roles inside the Service (organisation admin, pod lead, member, viewer) carry real permissions; granting a role is your decision, and actions taken under a role you granted are attributed to your organisation.
- You must be a business. The Service is offered to organisations, not consumers, and nothing in it is directed at individuals acting outside a business.
3. Connected tools
The Service works by connecting to tools your organisation already uses (Slack, Google Workspace, and others). For each connection:
- You confirm your organisation is entitled to grant the access it authorises — that connecting a source does not breach that provider's terms or someone else's rights.
- The scope of access is what you authorise, per connection; you can revoke a connection at any time and the Service stops reading from it.
- Slack specifically: installing the Mosaic Slack app is your organisation's explicit authorisation for the Service to use, process, and durably store the Slack API data the coworker can access — messages in channels it is a member of and direct messages sent to it, plus the workspace, channel, and user profile information needed to attribute those messages and operate the app — as the coworker's memory and as source material for your knowledge base. Uninstalling the app or revoking its tokens ends that authorisation, and the stored Slack API data is then deleted on the timeline the Privacy Policy states.
- Third-party platforms have their own terms, and your use of them through Mosaic remains subject to those terms. If a platform withdraws or restricts API access, the parts of the Service that depend on it degrade accordingly; that is outside our control, though we will tell you when it happens.
4. Your content and your knowledge
- You own your content. Everything your organisation puts into the Service, everything it authorises the Service to read, and the knowledge base built from it ("Customer Content") remains your organisation's. We claim no ownership.
- Our licence is to operate the Service, full stop. You grant us the rights needed to host, process, transmit, and display Customer Content in order to provide the Service, and no other rights. We do not use Customer Content to train AI models, and our AI providers are contractually restrained from doing so, as the Privacy Policy sets out.
- You can leave with it. Your knowledge base is yours; on request when you leave, we will assist you in exporting it in a portable form (markdown), and we then delete your tenant's data as the Privacy Policy describes.
5. The AI coworker — what you are agreeing to
This section exists because an AI coworker is not a spreadsheet, and the honest allocation of responsibility matters.
- AI output is generated, not warranted. The Service uses large language models. Their output can be wrong, incomplete, or confidently misleading, even with the citation, review, and grounding controls the Service applies. AI output is an input to your organisation's judgement, not a substitute for it. You are responsible for decisions and actions your organisation takes on the basis of AI output.
- Three kinds of AI action. The Service distinguishes: (a) answers and drafts — content produced for a person to read, which commits you to nothing; (b) governed proposals — changes to your knowledge base, which take effect only when a person your organisation authorised approves them; and (c) authorised actions — things the coworker does in connected tools or on a schedule because your organisation configured it to (a skill, a scheduled job, an approved automation). Category (c) actions are your organisation's actions, made with your authority; configure them with the same care as any delegation to a member of staff.
- Human gates are load-bearing. Do not build processes that bypass the approval gates the Service provides. If you configure the coworker to act without review where the Service offers review, the consequences of that configuration are yours.
- Who runs the models, and where. By accepting these terms you agree to Customer Content being sent, per-request, to every model-processing host Mosaic routes to, so that any model on the roster can run your work without our coming back to ask you again. Model processing takes place on hosts in the United States, except Google Cloud, which serves from several regions and is not pinned to one country. The companies that process your content to run the models are Anthropic, Baseten, DeepInfra, Fireworks AI, Google Cloud, OpenAI, SiliconFlow and xAI. None of them trains on your content. None retains it after the request, with narrower exceptions: Anthropic holds it for 30 days, reviewed by a human only if a request is flagged; OpenAI holds it for 30 days, reviewed by authorised staff only if a request is flagged. Our subprocessor page carries the full per-host record and is updated as the roster changes; material changes are notified as Section 11 describes. Which models your organisation's coworker uses remains your administrator's choice, made on the Models page in-product — but that choice is between hosts you have already been told about here, not a second consent.
- The models change. We may update the AI models used to provide the Service as the field moves; material changes to how Customer Content is processed are reflected on our subprocessor page and notified as Section 11 describes.
- Professional advice. AI output is not legal, financial, medical, or other professional advice, and no professional-client relationship is created by using the Service.
6. Fees and billing
- Fees are as stated at purchase (or in your order form), billed through Stripe. Prices are exclusive of GST, which will be added where applicable.
- Usage-based components (credits) are consumed as metered; your workspace shows consumption as it happens. Purchased credits are not refundable except as required by law, but survive for the period stated at purchase.
- If an invoice is unpaid 14 days after a reminder, we may suspend the workspace until it is paid. We will not delete Customer Content for non-payment without at least 30 days' notice.
- We may change pricing with at least 30 days' notice, effective at your next renewal - never mid-period.
7. Suspension and termination
- You can stop at any time: cancel renewal, export your data, and go.
- We may suspend a workspace immediately where necessary to address a security incident, a legal requirement, or a material breach of the Acceptable Use Policy — and will tell you why, and restore access when the cause is resolved unless termination follows.
- Either party may terminate for a material breach unremedied 30 days after written notice.
- On termination, Sections 4 (ownership), 8–10 (liability and disputes), and the export/deletion commitments survive.
8. What we promise, and the cap
- We warrant that we will provide the Service with reasonable skill and care and substantially as described. To the extent any statutory guarantee (including under the Australian Consumer Law) applies and cannot be excluded, nothing here excludes it; where the law allows us to limit our liability for breach of such a guarantee, we do so to re-supplying the Service or paying the cost of re-supply.
- Beyond what cannot be excluded: each party's total liability arising out of or in connection with the Service in any 12-month period is capped at the fees your organisation paid for the Service in that period, and neither party is liable for indirect or consequential loss, loss of profits, or loss of data (except to the extent caused by that party's breach of its confidentiality or data-protection obligations, or its wilful misconduct or gross negligence).
- These limits are mutual, and they do not apply to your payment obligations, to either party's infringement of the other's intellectual property, to either party's wilful misconduct or gross negligence, or to liability that cannot lawfully be limited.
9. Confidentiality
Each party will protect the other's confidential information with at least the care it applies to its own, use it only for purposes of this agreement, and disclose it only to people who need it and are bound to confidentiality. Customer Content is your confidential information. This clause does not prevent disclosures required by law, provided (where lawful) the disclosing party gives notice.
10. Disputes, law, and the rest
- These terms are governed by the laws of New South Wales, Australia, and the parties submit to the non-exclusive jurisdiction of its courts.
- Before litigating, the parties will attempt in good faith to resolve any dispute by escalation to senior representatives for 30 days.
- Neither party is liable for delay caused by events beyond its reasonable control.
- You may not assign this agreement without our consent (not to be unreasonably withheld); we may assign it as part of a corporate transaction with notice to you.
- If a clause is unenforceable it is severed; the rest stands. These terms plus any order form are the whole agreement; if they conflict, the order form wins.
11. Changes to these terms
We may update these terms. For material changes we will notify workspace administrators at least 30 days before they take effect; continuing to use the Service after the effective date is acceptance. Minor changes are simply published here as a new dated version. Each version is dated and prior versions remain available.
Contact: legal@mosaicailab.ai