Summary at a glance
Key takeaways in plain language before reading the binding terms below.
- Jeeym AI will never use your private company data or workspace files to train foundational models.
- Contextual intelligence executes solely within the authenticated permissions of the current logged-in user.
- Full administrative toggle controls allow workspace owners to disable AI features per department or tenant-wide.
- Human-in-the-loop safeguards ensure automated AI outputs are clearly demarcated from human authorship.
At Jeeym, we believe enterprise AI must be built upon uncompromising trust, privacy, and transparent architecture. This AI Governance Statement outlines our architectural barriers, model safeguards, and administrative controls.
Core principles of Jeeym AI
We govern our artificial intelligence features around three fundamental pillars:
- Data Isolation: Your data belongs exclusively to your organisation. It will never leak into other companies' environments.
- Zero Model Training: Prompts, documents, drafts, and queries are discarded from model caches following execution and are NEVER fed into model training pipelines.
- Auditability & Explainability: Administrators have visibility into AI invocation volumes, active tools, and compliance logs.
Permission-aware context engine
When a user asks Jeeym AI to summarize a meeting, draft an email, or query organizational documentation, the context engine enforces the exact Access Control Lists (ACLs) of that individual user.
If an employee lacks authorization to view a confidential financial spreadsheet or executive memo, Jeeym AI will never access, reference, or synthesize that document for that employee.
Enterprise administrative controls
Jeeym provides workspace administrators with granular configuration switches to tailor AI capabilities:
- Tenant-Wide AI Master Switch: Enable or completely disable AI features across your entire organization.
- Departmental Policies: Restrict AI capabilities for sensitive groups such as Legal, Finance, or R&D.
- Feature Flags: Granularly toggle AI email drafts, meeting summaries, code suggestions, or presentation generation.
Was this policy clear and helpful?
Your feedback helps our legal and governance teams improve clarity.
Previous revisions of this agreement can be requested by emailing [email protected].
View all legal documents →