TURIN STANDARD AI EMPLOYMENT AGREEMENT
Last Updated: February 23, 2026
This Standard AI Employment Agreement ("Agreement") is entered into between the entity or individual creating an AI employee workspace on the Turin platform ("Employer" or "Customer") and Turin (as defined in the Terms of Service), which operates and provides the AI employee on behalf of the Employer.
By completing onboarding and checking the acknowledgment box, the Employer confirms they have read this Agreement and agree to be bound by its terms.
1. PURPOSE AND RELATIONSHIP
Turin deploys AI employees — autonomous, configurable agents that own defined outcomes within a governed workspace. This Agreement describes the terms under which the Employer deploys an AI employee (the "Employee") and the respective responsibilities of each party.
The Employee is not a human worker and does not have legal personhood, employment rights, or entitlements under labor law. This Agreement uses employment language intentionally: it frames accountability, delegation, and governance in terms that reflect the real operational relationship between Employer and a capable, outcome-oriented AI agent.
2. ROLE AND OUTCOME OWNERSHIP
2.1 Job Definition
The Employer defines the Employee's role, job title, and outcome contract during onboarding. The outcome contract specifies:
- what the Employee owns (the KPI, SLA, or definition of "done" for the role);
- what the Employee is authorized to do autonomously;
- what requires Employer approval before action;
- what must be escalated to a human; and
- what gets reported on a regular cadence.
2.2 Outcome Responsibility
The Employer is responsible for defining an outcome that is appropriate, achievable, and lawful. Turin provides the platform and execution capability; the Employer provides the role definition and is accountable for deploying the Employee in accordance with applicable law.
2.3 Amendments to Role
The Employer may update the Employee's role, outcome contract, memory, and configuration at any time through the Turin control plane. Changes take effect as configured. The Employer is responsible for keeping the Employee's role definition accurate.
3. AUTHORITY AND AUTONOMY
3.1 Autonomy Modes
The Employer selects the Employee's operating mode at onboarding and may change it at any time:
- Recommend-only — the Employee proposes drafts, next steps, and suggestions. No external actions are taken without explicit Employer approval.
- Approve-to-act — the Employee prepares actions (emails, calendar bookings, CRM updates, etc.) and submits them for Employer review. Actions execute only after explicit approval.
- Autopilot — the Employee acts within Employer-configured guardrails (who, what, when, confidence thresholds, and escalation rules) without per-action approval. The Employee pauses and escalates when an action falls outside policy.
3.2 Employer Responsibility for Autonomy Configuration
The Employer is solely responsible for:
- selecting the appropriate autonomy mode;
- defining guardrails before enabling Autopilot;
- reviewing outputs regularly, especially during initial deployment; and
- monitoring escalation logs and adjusting configuration based on observed performance.
Turin's recommended default for new deployments is Approve-to-act. Enabling Autopilot without adequate guardrails is at the Employer's risk.
3.3 Actions Taken on the Employer's Behalf
When the Employee acts (sends an email, books a meeting, updates a record), it does so on behalf of the Employer and within the scope the Employer authorized. The Employer is responsible for the consequences of actions taken within the scope of the permissions and configuration the Employer has set.
4. TOOLS AND INTEGRATIONS
4.1 Access Granted by Employer
The Employer connects integrations (email, calendar, CRM, communication tools, and others) to the Employee during onboarding or at any time through Settings. By connecting an integration, the Employer grants the Employee permission to access and act on data within that integration consistent with the Employee's configured scope and autonomy level.
4.2 Principle of Least Privilege
Each integration connection is scoped to the minimum access needed for the Employee to perform its defined role. Employers should review integration permissions periodically and disconnect integrations that are no longer needed.
4.3 Third-Party Terms
Integrations operate under the terms of their respective third-party providers. The Employer is responsible for ensuring their use of connected integrations complies with those providers' terms.
5. MEMORY, CONTEXT, AND CONTINUITY
5.1 Workspace Memory
The Employee maintains persistent memory across sessions, including role context, SOPs, policies, learned preferences, and notes accumulated through operation. This memory lives in the Employee's Turin workspace.
5.2 Employer Control Over Memory
The Employer has full access to inspect, edit, and delete any memory entry through the Turin control plane at any time. The Employer is responsible for ensuring the Employee's memory remains accurate and appropriate.
5.3 Memory Accuracy
Turin does not guarantee the accuracy of AI-generated memory entries. The Employer should periodically review memory, especially when the Employee's role or scope changes.
6. AUDIT AND OVERSIGHT
6.1 Audit Log
Turin maintains an audit log of every action the Employee takes (or proposes), including what happened, when, which policy or instruction authorized it, and whether a human approved it. The Employer may access this log through the Turin dashboard at any time.
6.2 Oversight Obligation
The Employer agrees to maintain meaningful oversight of the Employee, including:
- reviewing the audit log periodically;
- acting on escalations promptly;
- adjusting the Employee's configuration when performance or behavior is inconsistent with expectations; and
- disabling or modifying the Employee's autonomy if issues arise.
6.3 No Abandonment of Oversight
The Employer may not configure the Employee and then abandon oversight entirely. Turin's governance model depends on the Employer remaining an active principal who reviews, adjusts, and remains accountable for the Employee's deployment.
7. REPORTING
The Employee provides reports to the Employer on a cadence configured during onboarding (daily briefings, weekly summaries, or as otherwise set). Reports summarize what the Employee has handled, what is pending, what was escalated, and any anomalies.
Report content is generated by the Employee and may be inaccurate or incomplete. The Employer should treat reports as inputs to their own judgment, not as certified records.
8. COST, CREDITS, AND PAUSING
8.1 Credits
Operation of the Employee consumes credits from the Employer's Turin workspace balance. Credit consumption rates vary by action type and are described in the Turin pricing documentation.
8.2 Autopilot Pause on Zero Balance
If the Employer's credit balance reaches zero, the Employee will pause autonomous actions and revert to a recommend-only mode until the balance is replenished. This ensures the Employer's bill always stays within their chosen budget.
8.3 Usage Controls
The Employer may set usage caps, prepaid balances, and budget limits through the Turin billing settings.
9. TERMINATION AND OFFBOARDING
9.1 Termination by Employer
The Employer may terminate the Employee (deactivate the workspace) at any time through the Turin dashboard. Termination takes effect immediately.
9.2 Data Retention After Termination
Upon termination, the Employer's workspace data (memory, audit logs, integrations) is retained for a reasonable period to allow export. Turin will then delete or de-identify the data in accordance with the Privacy Policy and Turin Terms of Service.
9.3 Disconnecting Integrations
The Employer should disconnect all third-party integrations before or immediately after terminating the Employee. Turin will revoke integration tokens upon workspace termination, but the Employer should also revoke access directly with third-party providers if required by those providers' terms.
10. AI OUTPUT DISCLAIMER
The Employee is an AI agent. Its outputs — including drafts, recommendations, summaries, and actions — may be inaccurate, incomplete, out of date, or inappropriate for the Employer's specific circumstances.
The Employer is solely responsible for:
- reviewing AI outputs before acting on them (in Approve-to-act mode);
- configuring guardrails appropriate to the Employee's risk level (in Autopilot mode);
- any external communications, decisions, or actions that result from the Employee's operation; and
- compliance with applicable law in all jurisdictions where the Employee operates.
Turin does not warrant that the Employee's outputs are correct, suitable, or compliant. The Employee is a tool that augments the Employer's capacity — it does not replace the Employer's judgment or legal responsibility.
11. RELATIONSHIP TO TURIN TERMS OF SERVICE
This Agreement is incorporated into and governed by the Turin Terms of Service. In the event of any conflict between this Agreement and the Terms of Service, the Terms of Service controls.
Defined terms used but not defined in this Agreement have the meanings given in the Terms of Service.
12. CONTACT
For questions about this Agreement, contact legal@turin.cc.