Acceptable Use Policy

Acceptable Use Policy

1. Purpose

This policy sets the rules for using the Graine AI platform. It protects the people your agents speak with, other customers, and the platform. It is the customer-facing policy; it is separate from Graine's internal policies for its own staff. It sits beside your contract, the Responsible AI Statement and the Shared Responsibility Matrix. If this policy and your contract conflict, the contract applies. Statements are true as of 3 Nov 2026.

2. Who is responsible

You decide why people are called or messaged and what your agents say. You are therefore responsible for how your agents are used, for what they say, and for everything done with your keys, members and numbers. Graine supplies controls to help; it does not take the duty away from you. The Shared Responsibility Matrix lists the customer-side controls.

3. Lawful use

1. You must use Graine only for lawful purposes and in line with the laws that apply to you and to the people you contact, including India's Digital Personal Data Protection Act, 2023 and TRAI's telecom commercial communication rules.
2. You must have a lawful basis for each use of personal data, give the notices the law requires, and obtain the consent the law requires, including for calls, recordings and transcripts.
3. You must not place unsolicited commercial calls or messages contrary to law. You must keep to the registration, number-series, do-not-disturb and calling-hour rules that apply to your calls, and check people against the registers that apply before you call. A workspace do-not-call list is Being rolled out; it does not replace your duty.
4. You must tell people they are recorded, and that they are speaking with an AI, wherever the law requires it. An AI-and-recording notice that an Admin can require for the whole workspace is Being rolled out and is off until an Admin sets it. Until it is live, your own greeting must carry the notice.

4. Honest and safe use of agents

5. You must not use an agent to deceive people about who or what they are speaking with where disclosure is required, or to impersonate a real person, company or authority without their consent.
6. You must not clone or imitate a voice without the speaker's consent. Use a voice clone only with a recording the speaker has agreed you may use for it.
7. You must not use Graine to harass, threaten, stalk or defraud anyone, to phish for passwords or payments, or to produce or spread illegal content.
8. You must not use an agent to make, or to materially drive, a decision with a legal or similarly significant effect on a person, such as one about credit, employment, insurance or essential services, without human review of that decision and a specific written agreement with Graine first.
9. You must not present an agent as a doctor, lawyer or other licensed professional. Agents do not diagnose health conditions, and the platform is not a medical device. You must not rely on an agent to reach emergency services, and you must give callers a route to a person.
10. You must review what your agents say before relying on it. Language models can be wrong.

5. Security and fair use

11. You must not try to breach the platform's security or the isolation between workspaces: no probing other workspaces' data, no use of another workspace's identifiers, no bypassing of rate limits, allow-lists or app and user checks, no malware, no tests of capacity without agreement.
12. You must keep secret API keys, passwords and webhook secrets secret, and tell Graine promptly at founders@graine.ai if you suspect they are exposed. You are responsible for activity under your keys.
13. You must not scrape the platform, reverse engineer its services or resell access beyond what your contract allows. The public SDKs are MIT-licensed and may be used under that licence, which does not extend to the platform's services.
14. You must stay within your rate and concurrency limits and not split work across workspaces or keys to get around them. If you need more capacity, ask Graine to raise your limits.
15. You must not use Graine in a way that harms other customers or the platform's availability.

6. Children

16. You must not use an agent to contact a person under 18 unless you have the verifiable consent of a parent or lawful guardian as the law requires (Digital Personal Data Protection Act s.9; Rules 10 to 12, from 13 May 2027). Graine does not check anyone's age; you must.

7. Consequences

17. Graine may ask you to change a use, give a warning, suspend access in whole or in part, or end your contract under its terms.
18. Graine may suspend access at once where there is a security threat or a legal order, and will tell you promptly.
19. Charges for conversations already held remain payable.
20. A breach of this policy by anyone using your workspace is your breach.

8. Reporting abuse

If you think a Graine-powered agent, a customer or a user has broken this policy, write to founders@graine.ai. Give the number or agent involved, the date and time, and what happened. A complaint about how your personal data was used can also follow the Grievance Redressal Policy. To report a security vulnerability, follow the Security Overview, and do not access, change or keep other people's data.

9. Changes

Graine may update this policy. The version in force is the one published at www.graine.ai/trust, with its version number and date, and the revision history below records each change.

Related documents

• Responsible AI Statement
• Shared Responsibility Matrix
• Security Overview
• Privacy Notice
• Grievance Redressal Policy

Revision history