Privacy Policy

About this policy

Tech Concierge by NewGen International — Privacy Policy. Version 1.34, effective 18 November 2026.

This policy applies to newgen.market, our blog at newgen.blog, the client portal at app.newgen.market, the Tech Concierge app for iOS and Android, the Telegram Mini App and the Concierge on WhatsApp and Telegram. Questions and requests: [email protected].

1. Who we are and what this policy covers

NewGen International Inc. ("NewGen", "we") is a corporation organised in the State of Florida, United States of America. With our affiliates we operate Tech Concierge, a technical department for small and medium businesses run by a named engineer together with an automated assistant. This policy explains how we collect and use personal data about the people who visit our websites, open an account, use the portal, the app or the Telegram Mini App, or write to the Concierge ("you").

The controller of your personal data is the NewGen entity that contracts, or would contract, with your organisation:

  • the United Arab Emirates and every country not listed below: NewGen International Inc., 1221 Brickell Ave, Suite 900, Miami, FL 33131, United States of America;

  • the European Economic Area, the United Kingdom and Switzerland: NewGen Europe OÜ, Tallinn, Estonia, our establishment in the European Union;

  • the United Arab Emirates, from the date announced on newgen.market: NewGen's affiliate in the United Arab Emirates.

Until your organisation adds its company details to its account, the controller is NewGen International Inc. We apply the standards of the EU General Data Protection Regulation (GDPR) and of the UAE Personal Data Protection Law (Federal Decree-Law No. 45 of 2021) to everyone, and the law of your own country where it gives you more.

If you are a customer of one of our clients: when a business uses a website, an app or a WhatsApp assistant that NewGen runs for it, that business is the controller of your data and NewGen is its processor, acting on its instructions under clause 10 and Schedule 4 of our Agreement with it. That business's privacy notice applies. Please contact the business; if we receive your request, we pass it on.

2. What we collect

  • Account and sign-in: your name, business email address, mobile number, role and language, and the languages your browser was set to when you opened the account. If you sign in with Google or Apple, the name and email address the provider shares with us and the identifier it gives us for you (Apple may give a private relay address instead of your own). One-time codes, which we keep only in scrambled (hashed) form; a password hash if you set a password; and sign-in sessions for your browser or device.

  • Free trial: when you open an account, the free trial that starts with it — its days, the work done and the allowance used. Until your organisation gives its details, its account carries your name. We ask for no trade licence, tax number or payment details during the trial.

  • Organisation and verification: your organisation's legal name, country, address, trade licence (the document you upload and the details on it), licence number and tax registration number (TRN), and the result of our review of the licence.

  • Contract and billing: the order form, the signed Agreement and the evidence of signature (the time, the IP address, the browser or device, and a fingerprint of the document), invoices, credit notes, payment status, the brand and last four digits of a saved card, and bank-transfer references. Card numbers are entered on Stripe's payment page and never reach us.

  • Using the service: requests, messages exchanged with the Concierge and our team (on WhatsApp, Telegram, email, the portal or the app), attachments, approvals, token ledger entries, disputes, satisfaction ratings and the videos you order.

  • Linked channels: your WhatsApp number or Telegram user ID when you link a channel, and the messages sent through it.

  • The app: a push-notification token, the platform, the app version and build, the operating-system version, the device model and your time zone (no advertising or device identifier), and the photos or files you choose to attach — the app asks for the camera or your photos only when you use them. Face ID and fingerprint unlock happen on your device; we receive only whether they succeeded.

  • Technical: IP address, browser and device type, and logs.

  • How our websites, the portal and the app are used (our own analytics, on our own servers): the pages and screens you open, the buttons and links you press, how far you scroll, the forms you start and send — never what you type in them — errors, how fast pages load, where a visit came from (the site that linked to us and a campaign's tags), and your country (read from your IP address, which we do not keep for this), device, system and browser, the language of the page and the first three languages your browser is set to. On newgen.market and on the portal's sign-in pages this is kept under an id your browser keeps only if you accept analytics, and then the visit may also be recorded — what the page showed and where you clicked and scrolled, with everything you type hidden; if you do not accept, visits are only counted, under a code that changes every day, and nothing is recorded. Once you are signed in to the portal or the app, your use is kept under your account, and a portal visit may be recorded with every text on the screen hidden; you can switch this off in Settings. When we learn who a visitor is — they sign in, write to us in the website chat, or open the plan we sent their business from a browser that accepted analytics — their earlier visits are linked to them.

  • Our blog, newgen.blog: our own analytics count its readers in the same way — the articles you open, how far down you read and for how long, the links you press, where your visit came from, and your country, device and browser, the language of the page and the first three languages your browser is set to. This is kept under an id your browser keeps for newgen.blog alone, and only if you accept analytics on the blog; if you do not, visits are only counted, under a code that changes every day. Visits to the blog are never recorded. A link from the blog to newgen.market or the portal carries the article's name in its address, so that we know which article brought a visitor; the message a WhatsApp button on an article prepares for you carries a reference to that article. If you go on to open an account, newgen.market and the portal pass that name on with your sign-up, and we keep it with the record of your sign-up, so that we know which article led to it; this needs no cookie. When a search engine, an AI assistant or another automated program reads the blog, the blog's server notes the program's name as the program gives it, the page, the time and the country the request came from: this counts programs, not people, and no IP address is kept.

  • Records of activity: to keep the platform secure and to be able to say who did what, we record every action taken in our systems — requests to the portal, the app, the API and our console with what was sent, sign-ins (including refused ones), sign-outs and password resets, each change to a record with its old and new values, the emails we send (address and subject) and our calls to outside services — with who acted, their role, the organisation, the time, the IP address, the browser or device and the app version. We never record passwords, access tokens, one-time codes, card details or signatures, and long values are cut short. Separately, a readable journal of activity (messages in and out with a short quote, requests, website chats, sign-ups, invoices and payments) lets our administrators follow what is happening. Only NewGen administrators can read either.

  • Website chat: if you use the assistant on newgen.market, the messages you write and its replies, and any name, email address, phone number and company you choose to give it. We do not keep your IP address, only a one-way scrambled (hashed) form of it, used to limit abuse.

  • Enquiries, referrals and outreach: what you send through the contact form on newgen.market (name, email, phone, company and message); the name and business details of a business a client refers to us; and, when we approach a business ourselves, the business contact details it publishes — on its own website or its Google Maps listing (such as its name, address, phone number, website, rating, and an owner's name, role and email), and, for a business we may offer our service to, the name, title, professional profile (such as LinkedIn), work email and mobile number of its owner or the person who runs it, which we look up by the business's website address in business-contact databases — the opening hours its Google Maps listing shows, how fast its website loads and the languages it is written in, whether it offers a WhatsApp button, a link to call it or an enquiry form, the links on its website to its listings on booking, tour and property platforms (such as Airbnb, Booking.com or Bayut) and to pages on social networks (such as Instagram or Facebook), its own or those of the people who run it, whether its website uses an online booking or payment system and which one, how long it took to answer a customer-style question we sent to its public number, whether that answer named a price and how many questions it asked back, whether and when its team wrote to us again after answering it, the messages we exchange with it, whether it opened the plan we prepared for it and which links in it were clicked, whether and when it opened the article on newgen.blog that one of our follow-up messages linked to (section 11), and our notes of that contact. When we approach a business we may build an AI assistant for it from what it publishes and from its answer to our customer-style question, and send it a link to a page where its owner can try that assistant: we keep when that page is opened and how many times, and what is written to the assistant there (section 11). A message may also come from a phone number we do not know to one of the WhatsApp numbers from which we send our customer-style question and our offers, and it may be from someone we never approached: we keep it, with that phone number, when it was sent and any photo, voice message or file it carries. Before any model reads it, our own rules compare its words and the name its WhatsApp shows with the names of the businesses we recently wrote to and of their owners: one that names a single business we wrote to from that WhatsApp number of ours joins that business's record, and one that may be from the owner of a business our offer reached goes to a member of our team; we do not keep that name.

  • LinkedIn: for the owner of a business we may offer our service to, or the person who runs it, whose LinkedIn profile we found as described above — the address of that profile, the invitation to connect we sent them, whether and when they accepted it, the messages we exchange there, and a picture of the part of our screen that shows each of these steps as our software takes it, or rehearses it without sending anything (their name, photo and headline as LinkedIn shows them, or our conversation). And for anyone who writes to that profile, or to our company page on LinkedIn (NewGen International Inc.), whether or not we wrote to them first: their name and headline as LinkedIn shows them, the address of their profile and the identifier LinkedIn gives it, and every message of that conversation, theirs and ours, with its time. Our software opens the conversations of that profile and of that page, and a person's profile only to learn its address; it opens no group conversations or sponsored messages, and takes nothing else from LinkedIn.

  • Instagram, Facebook and TikTok: when you comment on one of our videos or posts on Instagram, on our Facebook page or on TikTok, or write to us on Instagram, on Messenger or on TikTok, we receive through ManyChat (section 5) your username or name as the platform shows it and the identifier the platform or ManyChat gives your account, your comment and the video or post it is under, the messages you send us and the answers you choose in our replies, and our messages to you.

  • YouTube: when you comment on one of our videos on YouTube, we read through YouTube's API, with our own channel's permission, your public comment, the name and the identifier of your YouTube channel as YouTube shows them, the video it is under and when you wrote it, and we keep them with the reply we post under it (section 12).

  • Telegram: our team's Telegram account (Sam White, @samwhite_newgen) is connected to our servers as one more device of that account, through Telegram's own API. We keep the messages of its private chats and of the group chats it takes part in — those people send to it and those it sends — with when each was sent and edited and whether it was deleted in Telegram, the sender's name and username, and the phone number and username Telegram shows for the chat. What a message carries besides its text, such as a photo, a voice message or a document, is named by its kind, file name and size; the file itself is not kept. We do not keep the broadcast channels the account only follows, unless we turn them on (section 12).

  • WhatsApp: our team's WhatsApp account (Sam White, +971 52 950 7463), the number on which the Concierge answers our clients, is connected to our servers as a linked device of that account, in the way WhatsApp Web is. Besides clients' conversations with the Concierge (above), we keep a copy of every message on that account: those of its private chats and of the group chats it takes part in — those people send to it and those it sends, the Concierge's replies among them — with when each was sent and edited and whether it was deleted for everyone in WhatsApp, the sender's name as WhatsApp shows it and their phone number, and the name and number of the chat. What a message carries besides its text, such as a photo, a voice message or a document, is named by its kind, file name and size; the file itself is not kept. We do not keep statuses, channels or broadcast lists (section 12).

  • Threads: Our team's Threads account (Sam White, Tech Concierge by NewGen) looks for public Threads posts in which someone in the UAE asks for the kind of work we do. For a post we may answer we keep its text, its address and the author's public username, and an AI model (OpenAI) reads the text alone to tell whether we can help and to draft our public reply. A reply says who we are; it may say what we can do and link an article of ours about the same kind of work, and it gives no price. We answer a post at most once and never write to anyone privately first. A post we do not answer is deleted within 7 days. A post we answered keeps no username or text after 30 days, unless the author answers us: then their answer is kept as an enquiry, as this policy says of enquiries. Answer "stop", block our account or write to us, and we never write to you again on any channel.

  • Test drive: if your business takes our free test drive, what its customers write to the assistant we set up for it — their messages, names, numbers and enquiries. For that data we act as your business's processor, under the Test Drive Terms (newgen.market/test-drive-terms).

  • Sales partners: if you apply to be one of our sales partners or are one — your name, email address, phone number, company, country and city, what you tell us about your experience, the Partner Agreement you signed (the time, the address it was signed from and the name you typed), your certification of tax residence (your name, country, address and, if you give them, your date of birth and tax number), the businesses you register with us, and the commission and payouts on your statements. Your bank details and identity documents go to Stripe on its own pages and never reach us. If a sales partner told us about your business, the name, email address and phone number they gave us.

  • Waiting list: when we cannot yet serve your country, your business name and country, so that we can tell you when we can.

We do not knowingly collect special categories of personal data, such as health, biometric or religious data.

3. Why we use your data and on what basis

The lawful bases below are those of GDPR Article 6 and UAE PDPL Article 4:

  • Opening and running your account and its free trial before your organisation signs, including the set-up tasks and messages about them: steps taken at your request before entering into a contract.

  • Providing the service once your organisation has signed — the portal, the app, the Concierge, requests and deliverables: performance of a contract with you or your organisation.

  • Verifying a business and its signatory, including reviewing the trade licence and the TRN: legal obligations (know-your-customer, tax) and our legitimate interest in knowing who we contract with.

  • Billing, invoicing, tax records and collections: contract and legal obligations under tax and accounting law in the United States and, where it applies, the UAE.

  • Security, fraud and abuse prevention, and audit logs: legitimate interests in protecting the platform and our clients.

  • Records of activity and the journal of activity: legitimate interests in keeping the platform and our clients secure, preventing fraud, putting things right when they go wrong, and being accountable for who did what.

  • Improving the platform, using aggregated or pseudonymised data: legitimate interests.

  • Service messages — codes, invoices, incidents, approvals, renewals: contract, or steps before a contract.

  • Our own analytics on newgen.market, on newgen.blog and on the portal's sign-in pages, and Google Analytics: your consent, which you can withdraw at any time. Counting visits without cookies, which tells us nothing about who you are, and counting the automated programs that read our blog: our legitimate interest in knowing how our sites are used and found. Our own analytics of signed-in clients in the portal and the app, and linking a visitor's earlier visits once we know who they are: our legitimate interests in understanding how our service is used and how businesses come to us, to improve both; you can object at any time, and switch it off in Settings.

  • Marketing messages and case studies: your consent, which you can withdraw at any time.

  • Writing and checking the articles on newgen.blog from anonymised descriptions of work we did for a client (section 4): the consent the client's organisation gives with "NewGen may name my business as a client and show the work done for it", when it signs up or later in Settings, which it can withdraw at any time.

  • Answering you in the website chat, and following up when you leave your contact details: steps you ask us to take before a contract, and our legitimate interest in answering an enquiry you started.

  • Approaching businesses that may benefit from Tech Concierge — using the contact details they publish, measuring how quickly they answer a customer-style question and seeing whether they opened the plan we sent, the page of the AI assistant we built for them or the article of our blog we linked to (our team is told when that page is opened or its assistant is written to, and this may decide when and whether we next write to them and with which follow-up): for a business in the UAE, the PDPL's exception for data the person has made public (Article 4), and their consent once they reply; elsewhere, legitimate interests. A person at NewGen approves every message before it is sent or, when one of our administrators has switched on automatic approval, it is approved after the automated check described in section 4 and stays on record for a person to review; the messages we start, such as our customer-style question, our offers and our follow-ups, go on WhatsApp only on working days and by email on any day, weekends and public holidays included, between 09:00 and 18:00 UAE time (New York time for a business in the United States); our answers to a message you sent us may go at any hour of any day; and we stop the moment you say no or reply STOP — a refusal is final.

  • Finding the owner of a business we may offer our service to, or the person who runs it, and writing to them — one business offer and the conversation that follows — using the details we look up by the business's website address in business-contact databases: our legitimate interest in offering a business service to the person who decides on it, and their consent once they reply. A no or STOP from them or from the business stops all our messages to both: the owner's details are deleted, and both numbers are kept only in scrambled form on our do-not-contact list so that we never write again; our record of the messages exchanged is kept as the law requires.

  • As part of that one offer, inviting that owner, or the person who runs the business, to connect on LinkedIn from the profile of the member of our team who signs our messages and, once they accept, sending them one message — and no more unless they write back: our legitimate interest in offering a business service to the person who decides on it, and their consent once they reply. Our software sends the invitation and the message, within daily limits, and reads the replies; an invitation not accepted within 21 days is withdrawn. A no in any words, on LinkedIn or anywhere else, stops all our messages on every channel, as above.

  • Keeping the conversations people have with the LinkedIn profile of the member of our team who signs our messages, or with our company page there, as the record of our conversation with them, so that we can answer them and know what was said if they come back to us later: our legitimate interest in keeping a record of our own business correspondence. Recognising, with the AI described in section 4, whether someone who writes to us there may want our service, and adding them to our sales records so that a member of our team answers them: our legitimate interest in answering a business enquiry, and steps they ask us to take before a contract. We answer them only where they wrote to us and start no other messages to them, and a no in any words stops all our messages on every channel.

  • Keeping a message that comes from a phone number we do not know to one of the WhatsApp numbers from which we send our customer-style question and our offers, and telling, with the AI described in section 4, whether it is a business we asked answering from another phone, the owner or a manager of a business answering our offer from another phone, or someone we never approached, so that a business's answer counts as its answer and the rest is set aside: our legitimate interest in keeping a record of the messages that reach our own numbers and in measuring correctly how quickly the businesses we approach answer.

  • Answering a comment on our videos or posts on Instagram or on our Facebook page, or a message on Instagram, Messenger or TikTok, through ManyChat — one private message in reply to a comment, as Instagram's and Facebook's rules allow, and more only if you answer; on TikTok we answer only the messages you send us, never a comment — recognising, with the AI described in section 4, whether our service interests you, and adding you to our sales records if it does: our legitimate interest in answering you and in offering our service to someone who showed an interest in it, and your consent once you write back. Reply STOP and we send nothing more.

  • Answering a comment under our videos on YouTube, through YouTube's API, with one public reply from our channel — an answer from the article the video tells, written with the AI described in section 4, and where on our channel to find the full story; one reply in each thread, never a price or an offer — and leaving unanswered a comment of ours, one we already answered, praise or emoji alone, a link or spam: our legitimate interest in answering the people who ask about our videos. YouTube has no private messages, so our reply is public under your comment, as your comment is.

  • Keeping the messages of our team's Telegram account (section 2) — its conversations with clients, potential customers and partners — in one place with our other conversations, seeing each chat beside our record of the client or potential customer it belongs to (we tie the two by the phone number), and answering them: our legitimate interest in keeping a record of our own business correspondence and in answering the people who write to us, and, for a client, performance of the contract. No AI reads these messages (section 4).

  • Keeping every message of our team's WhatsApp account (section 2) — its conversations with clients, potential customers and partners — in one place with our other conversations, seeing each chat beside our record of the client or potential customer it belongs to (we tie the two by the phone number), and answering them: our legitimate interest in keeping a record of our own business correspondence and in answering the people who write to us, and, for a client, performance of the contract. Nothing is answered automatically from this copy, and no AI reads it (section 4); on that number the Concierge still answers only clients who have linked their WhatsApp number.

  • Looking for public Threads posts in which someone in the UAE asks for the kind of work we do, keeping a post we may answer and answering it once, publicly, from our team's Threads account (section 2), with the AI described in section 4: the PDPL's exception for data the person has made public (Article 4) and our legitimate interest in answering someone who asks in public for the kind of work we do, and their consent once they answer us. Answer "stop", block our account or write to us, and we never write to you again on any channel.

  • Running the sales partner programme — your application, the Partner Agreement, commission, statements and payouts: steps at your request before a contract, then performance of the contract; your tax certification and payout records: legal obligations. Contacting a business a sales partner registered with us, using the details the partner gave: legitimate interests. You can object at any time and we will stop.

  • Running a test drive for your business: steps at your request before a contract; for your customers' data, your instructions as controller (Test Drive Terms, clause 5).

  • Complying with the law and answering lawful requests: legal obligation.

4. The Concierge and artificial intelligence

  • The Concierge is an automated assistant. It classifies your requests, estimates their cost in tokens, answers questions from your account data and hands anything else to a named person. You can ask for a person at any time.

  • The artificial intelligence on the platform is a large language model provided by OpenAI, L.L.C. ("OpenAI"), in the United States, through OpenAI's API. It is used for every client, visitor and business we deal with, wherever they are. Two other providers have narrow roles: the coding agent our engineers use to do the work on your requests ("Aura", below) uses Google LLC's Gemini model, in the United States; and the articles on our blog are drafted with a model provided by Hangzhou DeepSeek Artificial Intelligence Co., Ltd. ("DeepSeek"), in the People's Republic of China, from texts from which every name and contact detail has been removed (below). Nothing that names or identifies you is sent to DeepSeek. For an organisation established in the United States, only the Concierge's replies use the model: nothing of it is translated, its licence is not read by the model, and nothing of it goes into the preparation and grading of our engineers' work or into our quality review.

  • When you write to the Concierge — in the app, the portal, on WhatsApp, Telegram or by email — we send OpenAI: what you write and up to ten earlier messages of your organisation's conversation with us, including your colleagues', the Concierge's and our team's; your name and role; your organisation's legal and trading names and country; its contract state, start date, monthly fee and currency, the name of its Named Engineer, its trial or plan, and the tokens it has left; its website domains, the project details in its Order Form and its standing instructions (up to 500 characters); and up to eight of its open requests, each with its number, title, state and date. When it needs them to answer, the Concierge also reads your organisation's token balance and ledger entries, its last five invoices (amounts, status and payment links), its statements, the terms of its contract, the details and estimates of its requests, the names of the businesses it referred to us and its knowledge notes. We never send passwords, one-time codes, card details, files, photos or voice recordings.

  • In the app and the portal we ask you, before your first message to the Concierge, whether the AI may be used with what you write. You can choose a person only, and change your choice at any time in Settings, or ask the Concierge for a person. If you choose a person only, or have not answered in the app or the portal, nothing about you or what you write is sent to the model: the Concierge does not use the AI and a person answers you; nothing is translated for you or about you, not even our own replies; your organisation's trade licence is read by a person; and your words are left out of the preparation, checks and grading of our engineers' work and out of our quality review. On WhatsApp, Telegram and email, where we have not asked you, the AI answers you, and what you write may be translated, until you choose a person only in Settings in the app or the portal or ask the Concierge for a person. Your choice covers the Concierge, translation and the reading of your licence. It does not cover the tools our engineers use to do the work you asked for: the coding agent and Google's Gemini model (below) work on your requests whatever you chose. The assistant for your customers and our blog follow their own rules, described below.

  • When your organisation uploads its trade licence and you have agreed to the AI, we send OpenAI the full text printed on it — not the file itself — which includes any names on it, so that it can read the company name, licence number, expiry date and business activities; a person at NewGen checks the result against your application before we rely on it. The licence of an organisation established in the United States is never sent.

  • Translation is done with OpenAI, only for people who agreed to the AI, and never for an organisation established in the United States. When someone at NewGen reads another language, we send OpenAI the text of what you write — chat and group-chat messages, the titles, descriptions and acceptance criteria of your requests, photo captions, the reasons for a dispute and the goals of a video — and, to show you our texts in your language, our replies, reports, notifications, invoice and ledger lines and the outcome of disputes. Only the text is sent, without your name or account details; we never send the Agreement, tax documents, statements, file names or code. When you write to us on WhatsApp, Telegram or by email in a different language from the member of our team answering you, their reply reaches you with a translation into your language above what they wrote. Messages are kept exactly as they were sent, with any translation beside them, and translations are stored so that the same text is not sent twice.

  • If we switch on an assistant that answers your own customers for you, we send OpenAI your business's name, website, greeting and rules, its knowledge entries and what your customers write to it. To draft its knowledge entries we also send the text of up to twelve pages of your website, your business's Google listing and any price list or catalogue you upload. This is part of the service your organisation asked for, and it does not depend on the choice described above.

  • Our engineers do the work on your requests with a coding agent ("Aura") on our own computer. So that it can do the work, we send Google's Gemini model the text of your request, your organisation's name, its project details and links to the files attached to the request; Google processes them in the United States. This is part of how we do the work you ask for, and it happens whatever you chose about the AI. To prepare the task for Aura, check that it is still working and grade the work at night, we also send OpenAI the request's title, description and acceptance criteria, your organisation's name, the names of the attached files, the project details, the text on Aura's screen, Aura's conversation and the rating you gave the work — only the words of people who agreed to the AI, and nothing of an organisation established in the United States.

  • At night we use OpenAI to review the quality of our engineers' work and the progress of each project. We send it the titles of requests, our engineers' notes, the project's description and goals, and the messages in your organisation's conversation with us in the period, from the people who agreed to the AI only. Nothing of an organisation whose owner chose a person only, or of an organisation established in the United States, is sent.

  • The assistant in the chat on newgen.market works the same way for visitors without an account: we send the language model your messages in that chat, up to the last ten, and any contact details you gave in it, so that it can reply. If you ask for a person, a member of our team reads the chat and answers you.

  • When a business we approached replies, OpenAI reads the reply to sort it and drafts our answer. Before any message to a business we approach is sent, the first one included, OpenAI may also check the draft. For that check we send it the draft, the business's details from section 2 (its name, trade, area and website, and what we measured, such as how quickly it answers and how fast its website loads), up to the last ten messages with the business, what is already waiting to be sent to it, and the terms of our offer. We do not send it the details we found about an owner or anyone else at the business, such as their mobile number or email; the draft and the messages themselves may contain the name of the person we are writing to. A member of our team reads, corrects and approves each message before it is sent, unless one of our administrators has switched on automatic approval: then a message the check passes is approved without a person and sent under the same rules, and one it does not pass waits for a person. Every message, the result of the check and who or what approved it are recorded, and our team can review them at any time. Reply STOP, or tell us no in any words, and we stop all messages to you.

  • OpenAI may also choose which of the facts we measured about a business we approach our offer to it names, and which one a follow-up may name: for example, that its Google Maps listing shows it closing early, or that its website sends guests to a booking platform and the pages we read show no way to book directly. For this we send it facts about the business alone: its line of work, its country and the facts our own rules found, each with what we measured (section 2), such as its closing time, how many minutes it took to answer our question or how fast its website loads. We send it no name, of the business or of anyone at it, no phone number, email address or web address, and no message. It may choose only among the facts our rules found, and it writes none of the message: every sentence that states a fact is written from our own template, never by the model, and the message is approved as the paragraph above describes.

  • Our invitations and messages on LinkedIn are drafted and checked by OpenAI, and approved, in the same way as our other messages to businesses we approach (two paragraphs above), from the same details of the business. To draft them we send it neither the person's name nor the address of their LinkedIn profile: our own systems add their first name to the text afterwards. What they write back is read and sorted, and our answer drafted, like any other reply, and those messages may contain their name. The pictures of our screen are never sent to a model. When someone who is not yet in our sales records writes to that profile or to our company page on LinkedIn, OpenAI reads their conversation with us — their messages and ours, and the headline LinkedIn shows for them — to tell whether they may want our service, are offering us their own, are looking for a job, are recruiting or are writing personally. We send it neither their name nor the address of their profile; the messages themselves may contain names.

  • When a message comes from a phone number we do not know to one of the WhatsApp numbers from which we send our customer-style question and our offers (section 2), OpenAI reads it, with the other messages waiting from that number, to tell whose it is: a business we asked our question, answering from another phone; the owner or a manager of a business answering our offer from another phone; or someone we never approached. We send it the messages' words and when they were sent (of a photo, voice message or file, only what kind it is), and the businesses we recently wrote to from that WhatsApp number of ours: the name, line of work and area of each, the question we asked it and when, whether and how quickly its own number answered, when our offer went to it and whether to the business or to its owner, and whether the messages quote our question to it or carry one of its numbers. We send it neither the number the messages came from nor the name on its WhatsApp, nor the words of our offer, nor the details we found about an owner or anyone else at a business, such as their name, mobile number or email; the messages themselves may contain names and numbers. Only where it is sure, and our own rules do not disagree, is its reading acted on without a person: the messages we hold from that phone number that nobody matched join that business's record, with the number, the first of them after our question as its answer to it, or the messages it read are set aside as nobody's; and where nothing but its reading ties them to that business, a member of our team confirms how quickly the business answered before any message of ours states it. Otherwise a member of our team decides, and always where OpenAI reads a message as an owner answering our offer or where one of our team is handling that business. We keep each message and what was read of it (section 7).

  • When you comment on our videos or posts on Instagram or on our Facebook page, or write to us on Instagram, Messenger or TikTok, our assistant answers you on our behalf: OpenAI reads what you write to us, with the earlier messages of that conversation and any contact details you gave us in it, to answer you and to tell whether our service interests you. So that the assistant can address you, OpenAI may also be given your name and, on Instagram, your username, as the platform shows them, but never the identifier of your account; the messages themselves may contain your name.

  • When you comment on our videos on YouTube, OpenAI reads your comment, with the public text of the article the video tells, to write our one public reply; we send it neither your name nor the identifier of your channel.

  • The messages of our team's Telegram account (section 2) are not the Concierge's: no AI model reads them. They are sent neither to OpenAI nor to DeepSeek nor to any other model, and nothing of them is translated.

  • Our copy of the messages of our team's WhatsApp account (section 2) is read by no AI model: it is sent neither to OpenAI nor to DeepSeek nor to any other model, and nothing of it is translated. A linked client's conversation with the Concierge on that number is handled, as before, as the paragraphs above say.

  • When our team's Threads account finds a public post on Threads (section 2), OpenAI reads the post's text alone, with a list of the work we do and, when we link one, the public text of an article of ours, to tell whether we can help and to draft our one public reply; we send it neither the author's username nor the address of their post.

  • When a business we approached writes back, we may answer with a short voice note in the voice of a member of our team, who recorded it with their consent. ElevenLabs turns the note's text into speech in the United States; the text may contain the name of the person we write to and the business's name.

  • Our blog, newgen.blog, is written with DeepSeek's model, or with OpenAI's when DeepSeek is unavailable; neither receives a name or contact detail for it. DeepSeek drafts its articles from public news and from work we have finished for clients who allowed their work to be shown ("NewGen may name my business as a client and show the work done for it", in the sign-up or in Settings), and then checks each draft: its facts against what we gave it, and whether a reader could tell which client it is about. We use only the work of clients established outside the United States: the work of a client established in the United States is never used for the blog, and nothing of it is sent to DeepSeek. For a client's work we send DeepSeek an anonymised description of it: the kind of work, the country and the line of business in general terms, how the request reached us, how quickly it was answered and done and whether on time, and the rating the client gave; and the text of the request (its title, its description and what it had to achieve) and of our engineer's report, with the names, contact details, addresses, websites and reference numbers we hold for the client and its people, and any e-mail address, phone number, web address or reference number, taken out before anything is sent. If you have chosen a person only, nothing you asked for is used, and if your organisation's owner has, nothing of your organisation's is. Our own systems also check every article before it is published, and a person at NewGen reads a sample of them every day. Which client's work an article came from is recorded and seen only by our team; if your organisation switches the permission off, the articles drawn from its work are taken down within a day.

  • How the providers treat what we send. OpenAI keeps what reaches its API for up to 30 days to detect abuse, then deletes it, and does not use it to train its models. Our coding agent uses NewGen International's Google Workspace account, with Google AI Ultra for Business; under the Google Workspace terms Google does not use what Aura sends it to train its models or have it reviewed by people, and it keeps it under those terms and its Cloud Data Processing Addendum. DeepSeek receives only blog texts from which every name and contact detail has been removed; we have instructed it in writing not to use them to train or improve its models. ElevenLabs only turns the text of a voice note into speech (above).

  • OpenAI and Google receive the data in this section only to provide these functions for us, under written terms that require them to protect it with the same or equal protection as this policy provides: OpenAI under its Services Agreement and the data processing addendum that forms part of it; Google under the Google Workspace agreement and its Cloud Data Processing Addendum for NewGen International's Workspace account, which our coding agent uses. DeepSeek receives no personal data.

  • All the personal data we send to these providers is processed in the United States, where our own servers are too (section 6). We send each task only the data listed in this section, and never passwords, one-time codes, card details, files, photos or voice recordings; you can keep your own words away from the AI by choosing a person only.

  • Decisions with a legal or similarly significant effect on you or your organisation, such as suspending the service or ending a contract, are taken under the rules of the Agreement and reviewed by a person before they take effect, unless your organisation agreed to automatic execution in the Agreement.

  • Conversations with the Concierge are logged, and a sample is reviewed by our staff for quality. We do not use your conversations to build advertising profiles.

5. Who we share data with

Service providers that process personal data for us (sub-processors), what for, and where. Each processes it only to provide its service to us, under written terms that require it to protect the data as this policy does (section 4 for the AI providers):

  • DigitalOcean — our servers, database and backups in the United States (New York), and our file storage and a daily encrypted copy of the database in the United States (San Francisco); the file storage also holds the recordings of visits (section 2).

  • Cloudflare — delivery and protection of our websites and API — its global network.

  • Google — our business email (Google Workspace), sign-in with Google, push notifications on Android (Firebase Cloud Messaging), business listings from Google Maps (Google Places), YouTube, where our videos are and through whose API we read the comments on them and post our replies, and, only with your consent, website analytics (Google Analytics) — United States and its global network.

  • Apple — sign-in with Apple and push notifications on iOS — United States and its global network.

  • Stripe — card payments, and payouts to our sales partners (Stripe Connect), for which Stripe also checks the partner's identity — United States and European Union.

  • Meta — WhatsApp messaging, including our team's WhatsApp account, whose messages reach our servers through a linked device of that account (section 2), and Instagram, Facebook and Messenger, where our videos and posts and the comments and messages on them are, and Threads, where our team's Threads account looks for public posts and posts its replies (section 2) — its global network.

  • Telegram — Telegram messaging, the Telegram Mini App, and our team's Telegram account, whose messages reach our servers through Telegram's API (section 2) — its global network.

  • DeepSeek — drafting and checking the articles on our blog, from texts from which every name and contact detail has been removed; it receives no personal data (section 4) — People's Republic of China.

  • OpenAI — the language model behind the Concierge, the assistant on our website, our clients' assistants for their customers and the drafting of their knowledge, translation, the reading of trade licences, the preparation, checking and grading of our coding agent's work, the review of our work's quality, sorting, drafting and checking our messages to businesses we approach and choosing which of the facts we measured about them our offers name, reading the messages that come from phone numbers we do not know to the WhatsApp numbers we approach businesses from, to tell whose they are, recognising potential customers among the people who write to us on LinkedIn, and answering comments and messages on Instagram, Facebook, Messenger and TikTok, and comments on YouTube and public posts on Threads; and the drafting of our blog's articles when DeepSeek is unavailable (section 4) — United States.

  • Google (Gemini, Google LLC) — the model our engineers' coding agent uses to do the work on your requests: the text of the request, your organisation's name, its project details and links to the files attached (section 4); under our Google Workspace agreement and its data processing addendum, with no training on our data — United States.

  • ElevenLabs — the voice of our voice notes to businesses we approach, made from a recording of a member of our team's own voice (section 4) — United States.

  • ManyChat — replying, on our behalf, to comments on our videos and posts on Instagram and on our Facebook page, and to messages on Instagram, Messenger and TikTok: your username or name and identifier, your comment, your messages and the answers you choose (section 2) — United States.

  • Slack — internal alerts to our team about errors and website enquiries — United States.

We also share data with your organisation (its account owner can see the requests, approvals and activity under its account); with our professional advisers, auditors and insurers, under confidentiality; with authorities and regulators where the law requires it, including tax authorities in the United States and the United Arab Emirates and data protection authorities; and with a buyer or successor of our business, under the same obligations and with notice to you.

We do not sell personal data and we do not share it with advertising networks.

6. International transfers

Our servers, our database and our file storage are all in the United States. Our team works from the United States, the United Arab Emirates, Portugal and Estonia, and some of the providers in section 5 process data in other countries — including DeepSeek, in the People's Republic of China, which receives only blog texts with no personal data (section 4).

Where data leaves the European Economic Area or the United Kingdom, we rely on an adequacy decision where one exists — for the United States, the EU-US Data Privacy Framework and its UK extension, for providers certified under them — or on the European Commission's standard contractual clauses, or the UK equivalent, with providers that accept them. Where data leaves the United Arab Emirates we rely on Articles 22 and 23 of the UAE PDPL, including where a transfer is necessary to perform the contract you or your organisation asked for. You can ask us for details of the safeguards for any transfer.

7. How long we keep data

  • An account whose organisation never signs an Agreement: deleted after 12 months without activity, or straight away if you delete it in Settings.

  • Your organisation's data once it is a client: for the life of the Agreement; then erased from our live systems 30 days after it ends and from backups within 90 days, except the records below.

  • Signed Agreements, invoices, credit notes and payment records: 7 years after the end of the tax year they relate to, as tax law in the United States and the UAE requires.

  • Requests, messages, deliverables and ledger entries: for the life of the account, exported to your organisation on request, then erased as above.

  • One-time codes: 30 days. Sign-in sessions not used for 180 days are ended and deleted.

  • Technical logs: 14 days. Records of activity: in full for 24 months; after that what was sent and changed, the IP address and the device are removed, and the record of who did what and when is kept for the life of the service. The journal of activity: the quotes and email addresses in it are removed after 90 days. If you ask us to erase your data, your entries in both keep only a record that an action took place — its type, time and internal references — and nothing of what was sent or changed, from where, or your name. Backups: no more than 90 days.

  • Our own analytics, newgen.blog's included: the record of visits and of the pages, screens and clicks in them, 13 months after the visit, after which it is anonymised; recordings of visits, 30 days. The record of automated programs reading newgen.blog: 13 months, after which only the daily totals remain.

  • Enquiries sent through the contact form: 24 months.

  • Website chats in which you left no contact details: 90 days after the last message. Chats in which you left contact details, and those details: 24 months after the last message, like an enquiry through the contact form.

  • Businesses we approached that did not become clients, and the records of our messages to them, including whether they opened the plan, the page of the AI assistant we built for them or the article of our blog that we sent them: 12 months after the last contact. If you asked us not to contact you, after that we keep only a one-way scrambled (hashed) form of your number or email address, for as long as we approach businesses, so that we never contact you again.

  • Messages that came from a phone number we do not know to one of the WhatsApp numbers from which we approach businesses (section 2), and what was read of them (section 4): 90 days after each arrived, then erased. Once one is found to be a business's, it and the others from the same phone number are kept with that business's record, as above.

  • Details of the owner of a business, or the person who runs it, that we looked up in business-contact databases: 180 days after we obtained them, unless they reply to us, after which the conversation is kept as for businesses we approached, above. If they or the business asked us not to contact them, their details are deleted at once and we keep only the scrambled form described above.

  • LinkedIn: the address of a person's profile, whether they accepted our invitation, and our messages there are kept as the owner's details and the conversation above. The pictures of our screen: 90 days after they were taken, and only NewGen administrators can see them. If they or the business ask us not to contact them, their details and these pictures are deleted at once, and only a scrambled (hashed) form of their profile's address is kept, with the numbers and email addresses above, so that we never invite or write to them again. A conversation someone has with the profile of the member of our team who signs our messages, or with our company page on LinkedIn, while they are not in our sales records: kept while it goes on and for 24 months after its last message, then deleted. Once we add them to our sales records, it is kept with that record, as for businesses we approached, above, and for a client as the client's data. Anyone can ask us to delete our copy of their conversation at [email protected].

  • Instagram, Facebook and TikTok: your username or name, identifier, comments, messages and answers, 24 months after your last message, unless we add you to our sales records, after which they are kept as for businesses we approached, above. If you reply STOP, we send nothing more, and only a scrambled (hashed) form of your identifier on that platform is kept so that we never write to you again; you can ask us to delete the rest at [email protected].

  • YouTube: your comment, the name and identifier of your channel, and our reply, 30 days after we read your comment; after that we keep only whether the comment was answered or left, under which of our videos and in which language, with no name, identifier or words. You can delete your comment on YouTube at any time, and ask us to delete our copy sooner at [email protected].

  • Telegram: each message of our team's Telegram account, 24 months after it was sent, then deleted; a chat's name, username and phone number go once none of its messages is left. Until then, a message deleted in Telegram stays in our copy, marked as deleted. The text of the messages is kept encrypted, and only NewGen administrators can see them. You can ask us for a copy of your messages, or to delete them sooner, at [email protected] (section 8).

  • WhatsApp: each message of our team's WhatsApp account, 24 months after it was sent, then deleted from our copy; a chat's name and number go once none of its messages is left. Until then, a message deleted for everyone in WhatsApp stays in our copy, marked as deleted. The text of the messages is kept encrypted, and only NewGen administrators can see them. A client's conversation with the Concierge is kept, as before, with the client's messages, above. You can ask us for a copy of your messages, or to delete them sooner, at [email protected] (section 8).

  • Threads: a post we do not answer — its text, its address, the author's username and our draft — within 7 days after we kept it; a post we answered — the username, the text, the address and our reply — 30 days after our reply. After that we keep only what we did with the post and why, what kind of post it was, the work it asked for and its language, with no username, address or words, and, for a post we answered, its identifier on Threads, so that we never answer it twice. If the author answers us, their answer is kept as an enquiry in our sales records, as for businesses we approached, above. If you tell us to stop, we keep a scrambled (hashed) form of your username so that we never write to you again; you can ask us to delete the rest at [email protected].

  • A test drive that did not become a subscription: its data, including your customers' conversations, 60 days after it ends.

  • Sales partner applications we decline: 12 months. A sales partner's account, statements and payout records: for the life of the partnership, and payout records then for 7 years after the end of the tax year of the last payout, as tax law requires. The contact details of a business a partner registered that did not sign: 12 months after the registration lapsed.

  • The waiting list for a country we do not serve yet: until 30 days after we tell you we have opened there, and in any case no more than 182 days after you last updated your entry.

  • Marketing consent and preferences: until you withdraw them, and a record of the withdrawal for 3 years.

8. Your rights

Depending on where you are, under the GDPR, the UAE PDPL or the law of your state you have the right to:

  • access the personal data we hold about you and receive a copy;

  • have inaccurate data corrected and incomplete data completed;

  • have your data erased where it is no longer needed or was processed unlawfully;

  • restrict processing in certain circumstances;

  • receive the data you gave us in a portable format;

  • object to processing based on legitimate interests, and to direct marketing at any time;

  • withdraw consent at any time, without affecting processing before the withdrawal;

  • not be subject to a decision based solely on automated processing that has legal or similarly significant effects, except as the law permits;

  • complain to a data protection authority: in the European Economic Area, the authority of your country or the Estonian Data Protection Inspectorate (aki.ee), which supervises NewGen Europe OÜ; in the United Kingdom, the Information Commissioner's Office (ico.org.uk); in the United Arab Emirates, the UAE Data Office.

Residents of California and other US states with privacy laws: we do not sell personal data or share it for cross-context behavioural advertising, and you may exercise the rights above in the same way.

To exercise a right, write to [email protected]. You can delete your own account in Settings in the portal or the app (section 9). We answer within one month and may ask you to confirm your identity. Where a request concerns data we process for one of our clients, we refer it to that client and help them answer it.

9. Deleting your account

You can delete your account in Settings in the portal or the app. If your organisation has not signed an Agreement, the account is deleted straight away, together with its sign-in sessions, linked channels and Google or Apple sign-in links, and an organisation that never signed is closed. If your organisation has signed, deleting is handled as a request to erase your data: we deal with it within 5 business days and keep only what the law or the Agreement requires us to keep, such as invoices and the signed Agreement.

10. Security

We protect personal data with access controls, encryption in transit, private storage for documents such as licences, invoices and agreements, separation of each client's data, logging of privileged actions, regular backups and confidentiality obligations for our staff. If a personal data breach is likely to put you at risk, we notify the competent authority and, where required, you, without undue delay and within 72 hours of becoming aware of it.

11. Cookies and similar technologies

  • newgen.market keeps your language and your cookie choice in your browser's local storage, and the region you chose with the switch in its header (the United Arab Emirates or the United States) in a cookie (ng_region, one year). These are needed for the site to work and are always on. Unless you have chosen a region, the site shows the version for the country your IP address is in, as Cloudflare reports it; it uses that country only to choose the version and does not store it. The first time you open newgen.market's home page, if the first language your browser is set to is one the site has other than English, the home page opens in it (in Spanish, the version for the United States if that is your region). The language it opened in, or the one you then choose with the switch in its header or the offer at the top of a page, is kept in a cookie (ng_lang, one year), so that this happens only once and your choice holds; like the region's, it is needed for the site to honour your choice. Every other page, and a search engine or another automated program on any page, is shown in the language its address names.

  • The chat on newgen.market keeps the id of your conversation, whether you have left your contact details and whether you have turned its sound off, in your browser's local storage, so that you can pick up the same conversation later. Clear this site's data to start afresh.

  • If you come to newgen.market through a sales partner's link, the site keeps that partner's code in your browser's local storage (newgen.partner), so that the partner is recognised if you sign up. Clear this site's data to remove it. In the same way, the campaign tags in the address you arrived at (utm_source and the like, which name the blog article that brought you) are kept for your visit in your browser's session storage (newgen.attribution), and in its local storage for up to 30 days if you have accepted analytics; the portal keeps them for the visit until you sign in, so that they go with your sign-up.

  • The client portal keeps your sign-in session and your language in your browser's local storage. Until you choose a language, it shows the first of your browser's languages that it has, or your account's once you sign in. These are needed for the portal to work.

  • Our own analytics, on newgen.market and in the portal: if you accept analytics, or once you are signed in to the portal, the site keeps an id for your browser (ng_vid, 13 months) and one for your visit (ng_sid, 30 minutes) in first-party cookies on newgen.market and its subdomains. If you decline or make no choice, neither is set and your visits are only counted. Our own team's browsers keep a cookie (ng_staff) so that their visits are left out.

  • Google Analytics runs on newgen.market and in the portal only if you accept analytics in the cookie banner. If you decline or make no choice, it does not load and sets no cookies. You can change your choice at any time with "Cookie settings" at the foot of every page of newgen.market.

  • If your browser sends a Global Privacy Control signal, newgen.market treats it as a refusal of analytics: it shows no cookie banner and loads neither Google Analytics nor our own analytics, so no ng_vid or ng_sid is set and your visits are only counted, as when you decline.

  • newgen.blog, our blog, keeps your cookie choice in your browser's local storage and asks for it in its own banner: a choice made on newgen.market does not carry over. If you accept analytics there, it keeps an id for your browser (ng_vid, 13 months) and one for your visit (ng_sid, 30 minutes) in first-party cookies on newgen.blog alone, which neither newgen.market nor any other site can read. If you decline or make no choice, neither is set and your visits are only counted. Our own team's browsers keep a cookie there too (ng_staff), so that their visits are left out. A Global Privacy Control signal is a refusal there too: no banner is shown and no cookie is set. The blog loads neither Google Analytics nor the Telegram script. You can change your choice at any time with "Cookie settings" at the foot of every page of the blog. The blog's home page, opened for the first time, opens in the first language your browser is set to if the blog has it and it is not English, and the blog keeps the language it opened in, or the one you choose with its language switch, in a cookie on newgen.blog alone (ng_lang, one year).

  • Cloudflare, which protects our websites, may set a short-lived security cookie (__cf_bm) to tell people from automated traffic. It is needed for that protection to work.

  • To support the Telegram Mini App, our websites load a script from Telegram, which lets Telegram see your IP address.

  • The app uses no cookies. It keeps your sign-in in your device's secure storage (the iOS Keychain or the Android Keystore) and, unless you switch usage analytics off in Settings, an id it makes for itself when installed, with which it sends the screens you open and the buttons you tap.

  • When we send a business its plan, the link is its own: our page records when it is opened and which links in it are clicked, with a cookie for that link alone and without any third party. If the browser opening it has accepted analytics on newgen.market, its visits to our site are linked to that business.

  • When we send a business the link to the page of the AI assistant we built for it (section 2), the link is its own: our server records when that page is opened and how many times, and sets no cookie for it; link previews and other automated programs we recognise are not counted.

  • Our second follow-up on WhatsApp, and our second and third emails, to a business that has not answered our offer may carry a link to one article on newgen.blog about its line of business. The link is that business's own: it leads through our own server — at newgen.market, or at the domain our email came from — which notes that the link was opened, which article it was, which of our messages carried it and when, and then sends the browser on to the article. Our server sets no cookie for it and keeps neither the IP address nor anything about the browser; link previews and other automated programs are not counted. The article's address carries only the campaign's tags — where the link was and which article — and nothing that identifies the business, so newgen.blog counts the visit as it counts any other (section 2).

  • We do not use advertising or cross-site tracking cookies.

12. Messaging platforms

When you write to us on WhatsApp or Telegram, the platform processes your data under its own terms and privacy policy, and we receive your identifier and the messages you send. We message you there only about your account and, outside WhatsApp's free-form window, only with templates the platform has approved. You can unlink a channel at any time on the Team screen.

When we invite the owner of a business to connect on LinkedIn or write to them there, and when someone writes to the profile of the member of our team who signs our messages or to our company page on LinkedIn, LinkedIn processes the invitation and the messages under its own User Agreement and Privacy Policy, and we keep our copy of the conversation as section 7 says. They can ignore or decline the invitation, remove the connection or block us on LinkedIn at any time, and a no, in any words, stops all our messages.

When you comment on our videos or posts on Instagram or on our Facebook page, or write to us on Instagram or Messenger, Meta processes your comment and messages under its own terms and privacy policy, and ManyChat passes them to us on our behalf. We reply to a comment with one private message, as Instagram's and Facebook's rules allow, and continue only if you answer. Reply STOP at any time and we send nothing more.

When you comment on our videos on TikTok or write to us there, TikTok processes your comment and messages under its own terms and privacy policy, and ManyChat passes them to us on our behalf. We do not answer a comment on TikTok with a message: we write to you there only in reply to a message you send us, within 48 hours of your last message and with no more than ten of our messages in that time, as TikTok's rules allow. Reply STOP at any time and we send nothing more.

When you comment on our videos on YouTube, Google processes your comment under the YouTube Terms of Service (youtube.com/t/terms) and Google's Privacy Policy (policies.google.com/privacy), and we read it through the YouTube API Services, with our own channel's permission. We answer a comment once, publicly, from our channel, and send no private message: YouTube has none. You can delete your comment or block our channel on YouTube at any time.

When you write to our team's Telegram account (Sam White, @samwhite_newgen), or it writes to you, in a private chat or in a group, Telegram processes the messages under its own Terms of Service (telegram.org/tos) and Privacy Policy (telegram.org/privacy), and we receive them through Telegram's API, as one more device of that account, and keep our copy as section 7 says. These are conversations with a member of our team, not with the Concierge, and no AI reads them. Deleting a message in Telegram does not delete our copy: ask us at [email protected]. You can block the account on Telegram at any time.

When you write to our team's WhatsApp account (Sam White, +971 52 950 7463), or it writes to you, in a private chat or in a group, WhatsApp processes the messages under its own terms and privacy policy, and we receive them through a linked device of that account, in the way WhatsApp Web is, and keep our copy as section 7 says. Our copy is kept apart from the Concierge: nothing is answered automatically from it and no AI reads it, and on that number the Concierge answers only clients who have linked their WhatsApp number, as above. Deleting a message for everyone in WhatsApp does not delete our copy: ask us at [email protected]. You can block the number on WhatsApp at any time.

When our team's Threads account (Sam White, Tech Concierge by NewGen) finds your public post on Threads or answers it, Meta processes your post and our reply under the terms and privacy policy of Threads, and we keep our copy as section 7 says. We answer a post at most once, publicly, under it, and write to nobody privately first. You can hide our reply or block our account on Threads at any time.

13. Children

Our services are for businesses and their staff. They are not directed at, and we do not knowingly collect data from, anyone under 18.

14. Changes to this policy

We publish any new version at newgen.market/privacy with its effective date and, for material changes, tell account holders by email at least 14 days before they take effect.

15. Contact

Privacy questions and requests: [email protected].

NewGen International Inc., 1221 Brickell Ave, Suite 900, Miami, FL 33131, United States of America. Establishment in the European Union: NewGen Europe OÜ, Tallinn, Estonia.

We keep only what the site needs in your browser to work: your choice of language. With your permission we also note how you use the site — the pages, what you click, and a recording of the visit with everything you type hidden — on our own servers, and count visits with Google Analytics. Without it we count visits without cookies and record nothing. Details are in our Privacy Policy.