Privacy policy

Using the NewsQuantified GPTs in ChatGPT

Our public GPTs — “Why Is My Stock Down Today?” and “Premarket Movers & Catalysts” — answer by calling our API at api.newsquantified.com. Each call sends us the query and nothing else: a ticker symbol and parameters such as the event class or the lookback window, plus the anonymous request identifier ChatGPT attaches (openai-ephemeral-user-id; where that is absent, a conversation identifier or the connecting IP address). We use that identifier for one purpose — enforcing the daily cap on the free demo, so one visitor cannot exhaust it for everyone.

We record each call — timestamp, which tool was called, whether it succeeded, and that anonymous identifier — to operate the service and measure demand. We do not receive your ChatGPT account, your name, your email address, or the rest of your conversation. Your use of ChatGPT itself is governed by OpenAI’s privacy policy, not this one.

Using this website

Pages carry no advertising and no advertising pixels. We use Google Analytics, but your browser never contacts Google and loads no Google code: the measurement runs on Cloudflare’s network, which sends the page view to Google from its servers instead. Be clear about what that does and does not change — Cloudflare passes your IP address on to Google as part of that report, because it is what makes country-level reporting work, so Google still learns your IP address and which page you were on. What changes is that no Google script runs in your browser. Our Content-Security-Policy will not load scripts from providers we have not named on this page, so adding one is a code change and this page is updated in the same change. Three things are recorded:

If you are in the EEA or the UK, none of the analytics below runs until you accept it. Google Analytics starts denied for visitors from those countries; the banner asks once and remembers your answer for six months. Declining is one click and the same size as accepting. Elsewhere, analytics runs by default and the banner lets you turn it off — but the campaign cookie described below waits for an explicit acceptance everywhere, which is stricter than the law asks of us outside Europe. Either way you can change your mind at any time through Cookie choices at the bottom of any page. Advertising signals are switched off for everyone, everywhere, and are never part of what the banner asks for.

• Arrival records — for each page request our own servers answer: the path, the address of the site that linked you here (without its query string), any utm_ or ref labels in the address, the country Cloudflare reports, whether the request looks like a crawler, and the answer you have given the banner. This one is not a cookie and does not wait for consent, because it records the request rather than the person: there is no identifier of any kind in it, no address, and nothing that can be joined to you or to another visit. Your banner answer is stored as one of the fields so that we can always tell these records apart from the ones you did agree to.
• First-party usage events — which page was viewed and which button was clicked, with obvious bot traffic discarded.
• Google Analytics — pages viewed and interactions on them, collected by Google on our behalf so we can see which pages and campaigns work. Cloudflare stores two first-party cookies for this on newsquantified.com: one named cfz_google-analytics_v4, holding a random identifier for up to a year so a returning visit can be told from a new one, and a companion that lasts only for the current visit. Neither can be read by scripts in the page. A third, zaraz-consent, records the answer you gave the banner and can be read by scripts. We have not turned on Google Signals or ads personalisation, so this is not used to build advertising audiences.
• Arrival attribution — a first-party cookie named nq_attrib, kept for 90 days and scoped to newsquantified.com, so a later sign-up is credited to the right source. It records the page you arrived on, the time you arrived, the address of the referring site if another site linked you here, and any utm_ or ref labels the link carried. The referring address is stored without its query string, so if you came from a search we keep the search engine, never what you typed into it. There is no identifier about you in it. A link carrying campaign labels replaces what is on file, because that is what the labels are for; an arrival without them is recorded only when the cookie is empty, so the way you first found us is never overwritten. It is a marketing cookie rather than a necessary one, so it follows the same rule the analytics above do. We show the banner where consent is legally required — the EU and the EEA, the United Kingdom — and there nothing is written until you answer it; decline, or simply ignore it, and neither cookie is ever written. Elsewhere both of these first-party cookies are written when you arrive, and Cookie choices at the bottom of any page turns them off at any time.
• Page activity — only where the analytics above is allowed for you, by the same rule: accepted in the banner, or, outside the EEA and the UK, not declined. Our own servers then record which pages you viewed on newsquantified.com, including the account portal and the documentation, which sections of a page came into view, clicks on marked buttons and on links that leave the page, how far down a page you scrolled, and how long the page was in active use. They also record the stages of a form you fill in — that you started it, that you submitted it, and the outcome — but never what you typed into it. With each page view we keep the path of the page, the campaign labels utm_source, utm_medium, utm_campaign and ref if the link carried them, and the size of your browser window only as a coarse class (phone, tablet or desktop) and a size rounded down to the nearest 100 pixels. The rest of the query string, the fragment (the part after #), any text from the page and anything you type are never recorded, and your IP address is not stored with it. The address of a site that linked you here is kept only as its host name, such as www.google.com. Two first-party cookies tie these together: nq_vid, a random identifier that lasts 180 days after your last visit so a returning visit can be told from a new one, and nq_sid, a random identifier for the current visit that expires 30 minutes after your last activity. Both are set by our servers by the same rule as the analytics above — outside the EEA and the UK unless you decline, inside them only once you accept — and neither can be read by scripts in the page. If you decline or withdraw consent, both are deleted the next time a page reports to us, and nothing further is recorded. When you sign in to the account portal, the nq_vid identifier in your browser is linked to your account.

If you submit your email address to receive reaction alerts, we email you a link and nothing is sent to that address until you click it. The address is kept from the moment you submit it as a contact with Resend, the provider that sends our email, and it is held there marked as unconfirmed and excluded from any mailing, so the one link is all it can receive. Confirming stores that address with the ticker or sector you asked about and the form you used, records it on the contact as confirmed, and we send you a small sample of the matching data as a spreadsheet.

If you accepted analytics, two things from the cookies above are stored on that contact as well: the random analytics identifier, and what nq_attrib holds about how you arrived — the campaign labels, the landing page and the referring site. That is how we can tell which pages, which campaigns and which other sites lead people to subscribe. Decline analytics and neither exists, so neither is stored — the contact then holds only your address and what you asked to hear about.

We use the address to send you what you asked for. We do not sell it and we do not pass it to advertisers. Every message we send includes an unsubscribe link; using it deletes the record and marks the contact as unsubscribed. If mail to your address permanently fails, or you report a message as spam, we stop sending to it and mark the contact the same way, without waiting to be asked.

If you ask us for an investor-relations readout or a call, the form also asks for your name, company, role, and anything you choose to type in the optional note. We store what you send with your ticker, and we use it only to prepare and hold that conversation. The same rules apply: not sold, not passed to advertisers, deleted on request.

Using the API and the MCP connector

An account is required. We store the account details you give our identity provider (email address, and organization name if you create one), your API keys, and a log of the requests those keys make — timestamp, tool, status, and volume — which we use for billing, rate limiting, and abuse prevention. Webhook delivery records are kept for 90 days and then deleted automatically.

What we never collect

We do not collect the content of your ChatGPT conversations, brokerage or trading credentials, portfolio holdings, payment card numbers (our payment processor handles those and we never see them), or any special-category personal data. Our GPTs are instructed never to ask for personal details in chat, and every Action they can call accepts only the query parameters listed in its published schema.

How long we keep things

• Demo and API request logs: 12 months.
• Arrival records: 90 days, then deleted by our analytics provider.
• Webhook event records: 90 days, pruned automatically.
• Arrival attribution cookie: 90 days, in your browser, clearable by you at any time.
• Page-activity cookies: nq_vid until 180 days after your last visit, nq_sid until 30 minutes after your last activity. In your browser, clearable by you at any time.
• Analytics cookies: up to one year for the identifier, the length of the visit for its companion. Your banner answer is stored twice — six months in our own cookie, which is what brings the banner back to ask again, and up to a year in Cloudflare’s. All in your browser, clearable by you at any time.
• Email addresses and investor-relations enquiry details you submit: until you unsubscribe or ask us to delete them.
• Account data: while the account exists, plus the period our tax and accounting obligations require.

Who else processes this data

We use a small set of infrastructure providers, each acting on our instructions: Cloudflare (website hosting and logs, the analytics measurement itself, and Turnstile — the check in front of our email forms that tells a person from an automated script; it sets no cookie of ours and we never learn who you are from it), Railway (API hosting), Clerk (accounts and authentication), Google (website analytics, which Cloudflare reports to on our behalf), Resend (email delivery — every message we send you passes through it, including the address it is sent to), and our payment processor. Data is processed in the United States and the European Union. We do not sell personal data, and we do not share it with data brokers or advertising networks.

Your choices

You can use the GPTs and this website without giving us your name or email address. Cookie choices, at the bottom of any page, reopens the banner wherever you are — including where you were never shown it — so withdrawing is the same two clicks as accepting, and takes effect immediately. You can also clear or block the attribution cookie in your browser, and unsubscribe from any email we send. One thing worth stating plainly, because it used to work and no longer does: tracker-blocking extensions and third-party-cookie blocking will not stop the analytics described above. Nothing loads from a third-party domain any more, and the cookies are first-party, so there is nothing for those tools to recognise. The Decline button, and clearing the cookies in your browser, are what stop it. Depending on where you live, you may also have the right to request a copy of the personal data we hold about you, to have it corrected, or to have it deleted — including data tied to an API account. Write to us and we will act on it; we answer these requests within 30 days.

Children

Our services are for professional and adult users. They are not directed at children under 16, and we do not knowingly collect data from them.

Changes

When this policy changes materially we update the date at the top of this page and, for account holders, send notice by email before the change takes effect.

Contact

Privacy questions and data requests: privacy@newsquantified.com. Account holders can also reach us from the dashboard at portal.newsquantified.com.

Nothing on this website or in our GPTs is investment advice. The data we publish describes what happened historically; it is not a recommendation to buy or sell any security.