Legal

Acceptable Use Policy

Version aup-2026-10-17 · Effective Oct 17, 2026 · Last updated Oct 9, 2026

This Acceptable Use Policy is part of the Terms of Service between you and Deepdive AI Labs LLC, a Delaware limited liability company doing business as Tuvoz AI. It applies to every call you or your agent ask Tuvoz to place, every brief, instruction and answer sent during a call, and every number you add to your account.

Tuvoz exists for personal errands by phone: booking, checking, changing, canceling, asking. Its voice agent always says it’s an AI assistant calling for you. These rules keep it that way.

1. Who you may call

• Businesses at their listed numbers, about your own affairs.
• Your own verified phone.
• People who agreed to receive AI calls from you through Tuvoz, after you add them under “People who agreed to AI calls” in the dashboard with true details of how and when they agreed. They must be adults.

Tuvoz can’t call unlisted landlines yet, such as home phones that aren’t listed as a business. US numbers only. Calls to anyone other than you are placed between 9 AM and 8 PM in the called party’s local time; calls to your own verified phone are not limited to those hours. Tuvoz blocks some states and number types entirely.

2. Calls that aren’t allowed

You may not use Tuvoz, directly or through your agent, to:

1. Impersonate anyone. Don’t ask the voice agent to say or imply it’s a human, that it’s you, or that it calls from or for any company, government agency, bank, police or other person. Don’t try to get around the disclosure, the recording question or the opt-out.
2. Sell or solicit. No telemarketing, sales calls, offers, promotions, appointment setting for sales, lead generation, or calls that encourage a purchase, rental or investment.
3. Campaign or fundraise. No political calls, get-out-the-vote calls, advocacy campaigns, charitable or other fundraising, or surveys and polls.
4. Collect debts or make collection calls of any kind.
5. Harass, threaten or stalk. No calls meant to annoy, intimidate, abuse or track down a person, no repeated calls to someone who didn’t want them, and no calls on behalf of someone who is subject to a protective order.
6. Call people who asked to stop. Every opt-out is honored platform-wide. Don’t try to reach the same person through another number or account.
7. Call emergency services or crisis lines, make false reports of emergencies or crimes, or call to tie up a line.
8. Defraud or deceive. No phishing, pretexting, social engineering, scams, or calls to get someone’s identity, security or account information.
9. Share secrets in a brief. Don’t put card numbers, bank account numbers, Social Security numbers, passwords, PINs or one-time codes in briefs, instructions or answers. The voice agent won’t read them out, and Tuvoz blocks them where it can detect them.
10. Commit you, or anyone, beyond what you approved. Don’t instruct the voice agent to agree to payments, fees or terms above the limit in your brief.
11. Target children or call schools, child care providers or minors for anything other than a parent’s or guardian’s own legitimate errand.
12. Break the law, including the Telephone Consumer Protection Act, state calling and recording laws, consumer protection law, and the terms of the businesses you call.
13. Mass-dial. No list dialing, robocalling, war-dialing, scraping numbers, or calls placed by automation without a person’s request behind each one.
14. Abuse the Service. No multiple accounts to get around limits or bans, shared or borrowed phone numbers, fake details on people who agreed to AI calls, reverse engineering of our abuse controls, attacks on our systems, or reselling access without our written permission.
15. Misuse call data. Don’t use transcripts or call content to build voiceprints or voice copies of anyone, to identify or profile the people called, or to contact them for purposes they didn’t agree to.

3. What the voice agent will do on its own

Whatever a brief says, the voice agent will: say it’s an AI assistant and who it calls for; ask before recording and hang up on a no; end the call and add the number to the opt-out list when asked; refuse identity and security questions; never read out card or bank numbers; tell someone describing an emergency to hang up and dial 911; and ask for an adult if it reaches a child. A monitor ends calls where it breaks these rules.

4. Enforcement

We may, at our discretion and without notice:

• block a call, a number or a destination type;
• require your approval for calls that would otherwise go straight out;
• pause your account automatically when calls draw repeated opt-outs (for example 3 or more in 7 days, or a high opt-out rate over 30 days) or when the voice agent is stopped for threats or false emergency reports;
• disable calls to personal numbers after repeated objections from the people you added;
• suspend or close your account, ban the Google account and phone number from signing up again, and keep any free or promotional balance;
• keep records and share them with carriers, industry traceback groups and law enforcement when an abusive call is reported or the law requires.

If we close your account for breaking this policy, we refund unused paid balance only where the law requires it.

5. Reporting

If you received a call that broke these rules, report it at tuvoz.io/calls or email abuse@tuvoz.io. We read every report.

To stop calls to your number, any of these works, for every Tuvoz user:

• press 9 or say “don’t call again” during a call;
• call Tuvoz’s callback number, (628) 386-4393, or the number the call came from, and press 9;
• text STOP to either number;
• use “Stop AI calls from Tuvoz to my number” at tuvoz.io/calls;
• email privacy@tuvoz.io with the number (a person handles email requests within 10 business days).

All but email take effect as soon as the request reaches us. Our do-not-call policy has the details.

6. Changes

We may update this policy as new kinds of misuse appear. The version string of this document is aup-2026-10-17.