Acceptable Use, AI & Communications Policy
A few terms are still being finalised. They are marked Not settled yet below. Where a decision has not been made we say so, rather than state a number we would not stand behind. If one of them matters to you, email johnsontechstudio@gmail.com and we will tell you where it stands, in writing, before you rely on it.
This document has not been reviewed by a lawyer yet. It is a first publication, drafted from an attorney-review package and grounded in what the software actually does, and the sections marked below are the ones still with our lawyer. If a section matters to a decision you are making, email johnsontechstudio@gmail.com and we will tell you where it stands, in writing, before you rely on it.
This policy sets the rules for using Johnson Tech Studio LLC dba gig33 (“Johnson Tech Studio,” “we,” “us”) services — your website, the AI receptionist and chat, follow-up messaging, and reviews. It is part of your Terms of Service. Where a phone carrier, Google, or another provider we depend on has stricter rules than these, theirs apply too.
It exists because we send messages and answer phones in your name, to your customers. When that goes wrong, the person on the other end does not distinguish between you and us, and neither does a carrier or a regulator.
1. Things you must never do with the service
- Anything illegal, fraudulent, deceptive, harassing, threatening, defamatory, discriminatory, or invasive of someone’s privacy.
- Impersonate a person, a government body, a professional, or another business — or hide the fact that a caller or visitor is talking to an AI.
- Get around our authentication, rate limits, do-not-contact lists, consent checks, or a carrier’s or platform’s rules.
- Use the assistant or a form to collect passwords, full card numbers, bank credentials, Social Security numbers, health records, biometric data, or other highly sensitive information, unless we have approved that use case in writing and put the right protections in place. The assistant is instructed to refuse; do not try to work around it.
- Present the service as an emergency line, or tell your customers an AI can stand in for 911, an ambulance, the police, or a licensed professional.
- Upload malware, scrape unlawfully, probe another customer’s data, or try to extract our system instructions, secrets, or another business’s information.
2. Using the AI honestly
- It has to say it is an AI. The assistant introduces itself as a virtual (AI) assistant and is instructed never to imply it is a person. You may not configure it to conceal that.
- It may not claim something happened until it has. The assistant will not tell a caller a message was taken or a transfer went through until the underlying action has actually completed. If it fails or times out, it must say so. You may not configure it to confirm bookings, payments, refunds, or anything else it cannot verify.
- Do not have it say things that are not true. No credentials you do not hold, no licences you do not have, no availability or staff you do not have, no guarantees you cannot keep, and no prices that are not your prices.
- Not for consequential decisions. Do not use the assistant to decide anything about someone’s employment, housing, credit, insurance, or medical care. It is a receptionist.
- Review what it does. AI misunderstands, mishears names and numbers, and sometimes states something wrong. Read your transcripts and leads. The consequences of acting on them are yours.
3. Phone calls and recording
- Inbound only. The assistant answers calls to your business. It does not call out. Someone calling you does not give anyone permission to have an AI call them.
- No outbound AI or recorded-voice calling. This is switched off and stays off. Not final — if we ever offer outbound AI or prerecorded calling, it will be a separate product with its own rules on consent, do-not-call lists, caller identification, quiet hours, and how someone withdraws permission — and our lawyer has to sign that off before it exists. Nothing on this page approves it.
- The recording announcement is not yours to remove. The assistant tells every caller, before anything else: “This call is recorded and transcribed, so we have a written copy.” It always plays. There is no setup option and no account setting that turns it off, and you may not configure, script, or instruct the assistant around it. If you think your situation genuinely requires otherwise, email johnsontechstudio@gmail.com and talk to us about it — that is a conversation with a person, not a setting you can change.
- What we actually do, stated exactly. The assistant announces the recording and then continues the call. It does not ask the caller to agree, it does not wait for an answer, and nothing stops the call if the caller objects — they would have to hang up. Not final — the stricter design our lawyer is reviewing would ask each caller for agreement and would not transcribe without one. That is not what the product does today. We are telling you the difference rather than letting the word “consent” cover it.
4. Text messages
- Text messaging is not live. Business texting needs a one-time carrier registration, and until it completes the system cannot send a text at all. Follow-up runs by email.
- When it is live, consent is per sender and per purpose. You need permission from the person, for messages from your business, about that subject. Permission to be called back about a quote is not permission to be sent promotions.
- Never bought, rented, scraped, or handed-on lists. Consent does not transfer with a spreadsheet.
- STOP works immediately. A reply of STOP suppresses that number in our system straight away, and every message is re-checked against that list at the moment it is about to be sent — not only when it was queued. A message already in the queue when someone opts out does not go.
- Opting back in takes a fresh, deliberate act by that person. We will not quietly resume.
- You are the sender. The messages go out in your name. The legal responsibility for having permission is yours.
5. Email
- Accurate sender details and honest subject lines. No disguising who a message is from.
- Commercial email carries a real postal address and a working unsubscribe, and we honour unsubscribes promptly.
- Do not pass an unsubscribed address to someone else to keep marketing to.
- Do not use a partner, an affiliate, or a vendor to do something you could not lawfully do yourself.
6. Reviews and reputation
This product is not switched on yet. These rules apply from the day it is.
- No fake reviews. Never write one, buy one, commission one, or use AI to invent a customer experience that did not happen.
- No sentiment gating. You may not decide who gets asked for a review based on whether you expect them to say something nice, and you may not offer anything of value in exchange for a positive one.
- No suppressing truthful negative reviews through threats or prohibited inducements.
- No insider reviews posted as if from an independent customer.
- Replies are drafts unless you deliberately turn on a setting that publishes them. Some reviews always stop and wait for you regardless of that setting — anything mentioning legal action, injury, discrimination or harassment, fraud or theft, a threat, a payment dispute, the police, or a serious safety problem.
7. Connected accounts
- Connect only accounts you own or are authorised to manage, and grant only the permissions the features you want actually need.
- Follow the rules of whatever you connect — Google, your calendar, your booking system, your phone carrier, your chat tool.
- We may reduce a feature, ask you to reauthorise, or disconnect an integration entirely if we have to in order to follow a provider’s rules or the law.
8. Data, and not keeping things forever
- Have us collect only what the job actually needs. Longer forms and chattier assistants mean more of your customers’ information sitting in a database.
- Do not treat “keep everything indefinitely” as the default just because storage is cheap. Ask us to delete what you no longer need.
- Act on deletion and correction requests from your own customers, and tell us so we can remove our copy.
- Do not export your customers’ details out of the service to sell, to market something unrelated, or to prospect for another business.
Not settled yet — we do not yet run a deletion schedule of our own. Our Privacy Policy explains exactly where that stands, including the routine that runs nightly and deliberately deletes nothing yet. Until it is live, “ask us and we will delete it” is the real mechanism, and it works.
9. Regulated industries
If you are in healthcare, dentistry, law, finance, insurance, childcare, or anything with licensing rules, your regulator’s requirements are yours to meet — including what your website claims and what your assistant may tell a caller. Tell us your limits and we will build them in. Do not put protected health information or other regulated data into the ordinary service without asking us first.
10. What we will do if this is broken
We may block a message, a call, a review reply, an integration, or an automation; ask you to put something right; switch off a feature temporarily; or end the service for a serious or repeated breach. We may keep evidence of what happened and cooperate with a lawful request from a carrier, a platform, or an authority. Where it is practical we will tell you first and give you a chance to fix it.
11. Reporting a problem
Abuse, a suspected privacy incident, an unlawful campaign, a compromised account, or anything on this page being broken — email johnsontechstudio@gmail.com, or write to us at Johnson Tech Studio LLC, 4319 Country Brook Dr, Dallas, TX 75287, United States. Not settled yet — we do not have a dedicated abuse or security mailbox yet. The address above is a real, monitored inbox and is the right place to send this until one exists.
Questions about this policy: johnsontechstudio@gmail.com