Who this is for: Users whose email is read by AI in HUBAL, and their Organisations. These Terms supplement the Terms, the Data Processing Agreement (DPA) and the B2B Annex; in matters of mail with AI they take precedence. The Polish version is binding.
Who accepts: the User — for themselves, when choosing a mode in which AI reads their mail; the owner or an administrator — also on behalf of the Organisation (when turning on the “Can use AI for mail” permission or when accepting for themselves). We record the acceptance with the document version, the person and the date.
§1. Scope
- Mail with AI means the Platform features in which AI analyses email from mailboxes connected to HUBAL: summaries, assignment to projects, suggested tasks and replies, commitments and deadlines, alerts, case details saved in the project journal and project card, knowledge search across email, and AI assistant answers based on email.
- The User chooses how much of their mail AI sees (Mail → Settings → “AI and your mail”): all mail, only the mail they choose, or none. A thread can be hidden from AI or revealed with the switch in its header, and a sender rule can make email from someone always or never visible to AI. The Organisation decides who may use mail with AI at all.
- Capitalised terms have the meaning given in the Terms.
§2. How AI works in mail
- AI works automatically, in the background, without a separate instruction from the User — for every email it may see under the settings.
- People in the Organisation see the outputs according to the permissions the Organisation has granted — e.g. journal entries, alerts and tasks are visible to people with access to the given project.
- The content of emails and the necessary context are sent to AI model providers — on the terms of §12 of the Terms and the Privacy Policy.
§3. Limitations of AI
- AI outputs may be wrong, incomplete, out of date or assigned to the wrong place (e.g. to another project or person).
- AI does not know what the Organisation or the User considers confidential. It may carry into its outputs information that should stay private or be known only to part of the team.
- Settings limiting AI (modes, hiding threads, sender rules, permissions) reduce this risk but do not guarantee that AI will recognise and leave out every piece of confidential information. Hiding a thread afterwards removes AI traces from that thread to the extent described in the user manual, but does not undo what someone has already seen or used.
- Mail with AI is an aid — it does not make decisions for the User or the Organisation.
§4. Obligations of the User
- Check AI outputs before you rely on them or pass them on; correct or delete wrong or unwanted entries.
- Hide private or confidential correspondence from AI (the switch in the thread, a sender rule) or choose “AI sees only the mail I choose” or “AI does not see my mail”.
- If in doubt, do not use mail with AI — you can choose “AI does not see my mail” at any time; the other mail features keep working.
§5. Obligations of the Organisation
- The Organisation decides whether and to whom it makes mail with AI available (the “Can use AI for mail” permission) and informs its Users and employees of the rules for using it.
- The Organisation ensures that the processing of correspondence by AI complies with the law — including the secrecy of correspondence, employment law (incl. Article 22³ of the Polish Labour Code) and data protection law — and that it has a legal basis for processing the data of the people it corresponds with.
- The Organisation ensures that AI outputs are reviewed before they are used or shared further within the Organisation, and decides who has access to the projects, alerts and tasks in which they may appear.
- The mail-with-AI settings in the Platform (the permission, modes, hiding threads, sender rules) are instructions of the Organisation as the controller within the meaning of §3(2) of the Data Processing Agreement (DPA).
§6. Liability
- Exclusions. Towards a Business (excluding ETCs), to the widest extent permitted by law, we are not liable for:
- the content, accuracy and completeness of mail-with-AI outputs;
- the disclosure in those outputs, to persons within the Organisation, of information that AI did not recognise as confidential;
- the way mail-with-AI outputs are used and decisions made on their basis;
- lost profits and indirect or consequential damage.
- Indemnity. A Business (excluding ETCs) shall indemnify us and cover damages and reasonable costs (including legal costs) if a third party — including an employee, associate or correspondent of the Organisation — or an authority brings a claim or proceedings against us arising from the processing of correspondence by mail with AI that the Organisation has enabled or allowed, or from a breach of §4–§5. The indemnity does not cover claims resulting from our breach of our obligations as a processor (GDPR, DPA). We will notify the Organisation of the claim and allow it to take part in the defence.
- The limitations and exclusions in this section do not apply to damage caused intentionally or where mandatory law does not permit them — in particular towards Consumers and ETCs and towards data subjects. Paragraphs 1–2 do not apply to Consumers and ETCs; by accepting these Terms, Consumers and ETCs confirm that they have been expressly informed of the limitations of AI described in §3 and knowingly accept them.
§7. Opting out and changes
- You can stop using mail with AI at any time by choosing “AI does not see my mail”; the Organisation — by turning off the “Can use AI for mail” permission. This applies to email arriving from that moment; what AI has already processed can be hidden from AI in the thread.
- We will present a new version of these Terms for acceptance in the Platform. Until it is accepted, the previously accepted version applies, and choosing a mode in which AI reads mail requires accepting the new version.
§8. Contact
Formal matters: legal@hubal-project.com · personal data: privacy@hubal-project.com · support: support@hubal-project.com. Service Provider details — Terms §19.