Anthropic: The $83 IPO, the $1.5 Trillion Debut — and What Claude Does With Your Email and Chats
Anthropic: The $83 IPO, the $1.5 Trillion Debut — and What Claude Does With Your Email and Chats
Anthropic is the most-searched thing in America right now. Not a celebrity, not a sports final — an AI company. Google Trends US has "anthropic" at number one with over 200,000 searches and a 700% spike, and the reason is simple: the company behind Claude went public today.
The numbers are the kind that make even people who never read a prospectus look up. Anthropic priced its IPO at $83 a share on November 18, above its own $70–$78 range, and sold 980 million shares — $81.34 billion raised, a valuation of roughly $1.5 trillion, the largest public debut on record. Trading under the ticker ANT began on the Nasdaq this morning, November 19. Its co-founders, Dario and Daniela Amodei, rang the opening bell.
That is the news. This post is about the part of the story that does not fit in a market headline: what happens to your email address and your conversations when you sign up for the product that just became the most valuable AI company on earth. We covered the roadshow and the first wave of fake "Buy ANT" emails on November 10 — this is the trading-day update: what actually priced, what changed since, and what Claude does with your data. I read Anthropic's privacy policy line by line, checked the company's own filing disclosures, and re-tested the Claude signup with a disposable address on IPO day. The IPO will be forgotten by next quarter. The data settings you pick this week will follow your chats for years.
The numbers behind the spike
Start with the figures, because they explain why this search term exploded and why it matters that you hear about the company from someone other than its bankers.
Anthropic sold 956,351,354 shares itself, with existing stockholders selling a further 23,648,646. The deal was upsized from 825 million shares after demand came in above the range. The underwriters hold a 30-day option to buy up to 147 million more shares, and the offering is expected to close on November 20. The previous record for a debut belonged to SpaceX, which raised $75 billion in June 2026 selling 556 million shares at $135 each. Anthropic cleared it by more than $6 billion.
A $1.5 trillion valuation puts a company founded in 2021 — by a group of ex-OpenAI researchers who left over safety disagreements — in the same conversation as the largest companies on the market, before it has ever reported a quarter as a public company. More than 400 Anthropic employees may be worth roughly $100 million each after the listing, according to CNBC's reporting. That is the scale of money now attached to a chatbot's answers.
None of that tells you whether Claude is safe to use. It tells you why everyone is suddenly searching the name, and why the next few weeks are the worst possible time to trust an email that mentions it.
Wait — what does Anthropic actually do?
Strip away the ticker and Anthropic is an AI research company whose only product most people touch is Claude, the chatbot that competes with ChatGPT and Gemini. The company was founded on a safety pitch: build the most capable models while publishing the guardrails that are supposed to keep them under control. That pitch is now a legal document, and this week it produced two stories that did not make the IPO headlines.
The first is in Anthropic's own S-1 filing. The company disclosed, with little fanfare, that it has activated part of its 2023 Responsible Scaling Policy — the internal rulebook that demands extra oversight and security once Anthropic's rate of progress crosses a defined line. The line is "compressing two years of AI progress into a single year," measured against the 2018–2024 pace. A report this month from METR, written with access to unreleased models, estimates Anthropic made roughly two years of historical progress in 12 months, with models improving about twice as fast as last year, largely because researchers now use AI tools to do the research. Anthropic's own responsible scaling officer, Sam McCandlish, said on social media that many of METR's conclusions match his own, while adding that the speed-up is hard to quantify. The prospectus itself warns that new models could "pose catastrophic or existential risks to humanity," and critics — including a coalition led by the Center for AI Safety, with founders of Hugging Face and xAI — have sharpened their tone, calling Anthropic "not a safe AI company" and accusing it of using "safety branding to court investors, employees, and the public."
The second story is a US Government Accountability Office report, GAO-27-108814, released Wednesday. Under covert testing in spring 2026, Claude and the three other most-used chatbots — ChatGPT, Gemini, and Grok — all produced workable attack plans once investigators got past their safeguards. Senator Maggie Hassan plans hearings on AI risks in January.
Neither story means you should delete Claude. Both mean the company now sitting at the top of the search charts is under more scrutiny, from more directions, than any AI company has faced — while also being worth more than any AI company has ever been at debut. That combination is exactly when a privacy policy stops being boilerplate and starts being the document that decides what the company can do with your words.
The signup: your email is the first thing Claude takes
Claude's signup is famously minimal. There is no username to invent and no password to create. You type an email address, Claude sends you a magic link, you click it, and you are in. We tested that flow in depth in October, when it worked end to end: the email inbox is the entire login, because there is no password to fall back on. This morning, that flow hit a wall before it even started — the details are in the re-test section below.
Anthropic's privacy policy, which I read in full on November 19, says what that minimal form actually collects. At signup it takes your name, email address, and phone number. If you pay, it takes payment information. And in certain circumstances it can ask you to verify your age or identity — in which case the data collected can include an image of a government-issued identity document, a photo or video of you, and facial geometry templates, which the policy notes may count as biometric data in some jurisdictions.
That last line is easy to miss and worth pausing on. A chatbot that most people think of as a text box can, in some verification flows, hold your face. The policy does not say every user is asked for ID. It says the company may ask, and that it collects that data when you choose to go through with it. If you ever get a verification prompt from Claude, you are not just confirming an email — you are deciding whether to hand a biometric template to a company whose shares started trading this morning.
The email address itself matters more than it looks, because of the magic-link design. Every future login — new device, new browser, expired session — sends another link to that same inbox. Lose the inbox and you have not lost a recovery option; you have lost the login. Our October test of a disposable address for Claude signup worked end to end, and it also showed the lockout that follows when a temporary inbox expires. A temp address is fine for trying Claude for an afternoon. It is a bad key to a paid account, a project full of your work, or anything you will want back in six months. Use a permanent address for the real account, and a disposable one only for the test drive — that is the split we recommend for every AI tool, not just this one. This morning's re-test, though, found the door closed for a different reason entirely — see the test below.
Your chats are training data until you say otherwise
This is the section most coverage of the IPO skips, and it is the one that matters most to anyone who actually uses Claude.
Anthropic's privacy policy says the company may use your Inputs and Outputs — everything you type, upload, paste, or ask Claude to do, and everything it returns — to train and improve its AI models, unless you opt out through your account settings. That is the default. Your chats are training material until you go into settings and tell the company otherwise.
There are two exceptions even after you opt out. If your conversations are flagged for safety review, Anthropic will still use them to improve its ability to detect harmful content, enforce its policies, or advance AI safety research. And if you explicitly report material through a feedback mechanism, that material can be used too. Rating an output with a thumbs up or thumbs down can store the entire related conversation as feedback. None of this is hidden — it is all in the policy — but none of it is shown to you when you are mid-conversation either.

Deleting a conversation helps, but on a schedule. The policy says a deleted conversation is removed from your history immediately and automatically deleted from Anthropic's back-end within 30 days. That is a real, stated number — and it is also a reminder that "delete" in a chat app means "queued for deletion within a month," not "gone now." Anything you would not want read back to you should not be typed in the first place; the delete button is a second line of defence, not a shredder.
The rest of the policy fills in the picture. Anthropic says it does not "sell" your personal data as the term is defined by privacy laws, and you can opt out of sharing for targeted advertising. Your data is transferred to Anthropic's servers in the US, and to other countries where its affiliates and processors operate. If you connect Claude to third-party services, Claude may send your Inputs and Outputs to those services, which then process them under their own privacy policies — a connection you enable is a data-sharing decision, even when it feels like a convenience. And in a merger, corporate transaction, or bankruptcy, your personal data can transfer as a business asset. A $1.5 trillion public company is not about to go bankrupt, but the clause exists, and the IPO is precisely the kind of corporate event the clause was written for.
My blunt read: the policy is more transparent than most, and still defaults to the company. Training on by default, opt-out buried in settings, safety-flag carve-outs, 30-day deletion, third-party sharing on connection. If you use Claude for anything sensitive — work documents, health questions, legal drafts, personal writing — the first thing to do after signup is open Privacy Settings and turn training use off. It takes thirty seconds and it is the single highest-leverage click in this whole story.
IPO day re-test: Claude signup with a temp email
I re-tested the signup this morning, November 20, between 00:31 and 00:39 IST, using a fresh disposable address generated on TempoMailUSA — testfix123@tempomailusa.com — and the result was not what October's test produced.
Getting to the signup page took some doing: claude.com redirected to claude.ai behind a Cloudflare "Verify you are human" check that reset itself twice before clearing on a fresh browser route. Once on the login page, I entered only the temp address and clicked "Continue with email." The form did not advance to a verification-code screen. Instead, directly under the email field, Claude returned this, verbatim: "Unfortunately, Claude is not available to new users right now. We're working hard to expand our availability soon."
No verification code was sent — the temp inbox stayed empty — and no phone number was requested. Crucially, nothing in the message named the disposable address or the temp domain. This reads as a general new-user block, not a rejection of temp mail: on the morning its maker went public, Claude appears to have closed the front door to everyone new, presumably under IPO-day demand. Our October test, when signup worked, is now out of date on this one point, and I will re-test once signups reopen rather than guess at the cause.
What this means for you is simple: if you were planning to try Claude today because it is trending, you may not be able to create an account at all — temp mail or not. Do not pay anyone offering to "get you in," and do not hand your permanent email to a third-party "Claude access" page promising a workaround. The door is closed at the source; wait for Anthropic to reopen it.
The scam emails this IPO will spawn
Every record-breaking IPO produces the same second wave: emails promising you a piece of it. Anthropic's debut is a perfect lure — a famous name, a ticker everyone is searching, a $1.5 trillion valuation, and millions of people who missed the offering and want in. Expect "ANT shares," "pre-IPO allocation," "claim your Anthropic stock," and "Claude Pro free for IPO week" messages within days, if they are not already in your spam folder.
To be precise about what I verified and what I did not: as of writing on November 19, I found no published report of a specific fake-Anthropic IPO email campaign. The warning below is not a claim that one exists. It is the pattern that followed every comparable debut, and the checks that catch it are the same ones that caught the fake settlement emails we documented this week.
The checks, in order:
- Anthropic never sells shares by email. Shares trade on the Nasdaq through a broker. Any message offering to sell you ANT directly, reserve an allocation, or take payment to "register" you is fake by definition.
- Check the sender domain, not the display name. Real Anthropic mail comes from anthropic.com. A display name reading "Anthropic IPO Team" on a lookalike domain is the oldest trick in the book.
- Never pay to claim something free. "Pay a fee to unlock your shares" and "verify your wallet to receive stock" are the two sentences that end every version of this scam.
- Treat urgency as evidence, not as a reason to click. "Allocation closes in 2 hours" is not a deadline; it is the pressure that makes people skip check one.
- Verify through the filing, not the email. Anything real about a public company is in its SEC filings and its own press releases. If the email's claim is not there, the email is the whole story.

The practical defence is the same one we give for every high-profile event: give new services a disposable address first. A temp inbox lets you try Claude, claim a beta, or sign up for IPO news without handing a company your permanent address on the day its name is at the top of every search chart. If the service turns out to be real and worth keeping, move the account to your permanent inbox while you still control both. If it turns out to be a lure, the address dies with the tab and the spam dies with it. That is what disposable email is for, and IPO week is when it earns its keep.
My call: the IPO is noise, the settings page is the story
Anthropic's $83 debut is genuinely historic, and it is also mostly irrelevant to you. You do not own the shares, you cannot buy them from an email, and the valuation will move on numbers that have nothing to do with your chats. What the IPO changes for an actual Claude user is narrower and more concrete: the company holding your conversations just became a public company worth $1.5 trillion, under a GAO report, a safety-policy trigger, and a Senate hearing in January — and its privacy policy still trains on your chats by default.
So here is the honest order of operations if you use Claude, or are about to try it because it is trending: sign up with a permanent address for anything you want to keep, and a disposable one for a test drive. Open Privacy Settings and opt out of training use. Treat the delete button as a 30-day queue, not an eraser. And delete any email offering you Anthropic shares without opening it — the real ones are at your broker, not in your inbox.
The market will decide what Anthropic is worth. You get to decide what your words are worth. Make that decision in the settings page, not in the headlines.
Frequently asked questions
What is Anthropic? Anthropic is an AI research company founded in 2021 by former OpenAI researchers, best known for Claude, its family of AI chatbots. It is a public benefit corporation that presents AI safety as a core part of its mission, and it went public on the Nasdaq on November 19, 2026, under the ticker ANT.
Why is Anthropic trending? Anthropic priced its IPO at $83 a share on November 18, 2026, raising $81.34 billion at a valuation of about $1.5 trillion — the largest IPO on record. Trading began on November 19, the same day a GAO report found Claude and three rival chatbots produced workable attack plans in covert testing, and the company disclosed it had activated part of its Responsible Scaling Policy.
What is Anthropic's stock price? The IPO was priced at $83 per share. That is the offer price, not the trading price — once shares begin changing hands on the Nasdaq, the price moves with the market, so check a live quote rather than trusting any article's number, including this one.
Does Claude train on my chats? Yes, by default. Anthropic's privacy policy says it may use your Inputs and Outputs to train and improve its models unless you opt out in your account settings. Conversations flagged for safety review and material you explicitly report can still be used even after you opt out.
Can I delete my Claude chats? Yes. Deleting a conversation removes it from your history immediately and from Anthropic's back-end within 30 days, according to the privacy policy. Deletion is not instant on the company's servers, so treat it as a 30-day process.
Can I sign up for Claude with a temp email? As of this morning (November 20), no one new can sign up at all — Claude returns "not available to new users right now" before any verification step, temp address or not. In October the same test worked end to end, and the magic-link design means the inbox is the entire login: there is no password, so if the temp address expires, you lose access to the account. Use temp mail for a test drive and a permanent address for anything you want to keep — once signups reopen.
Is the Anthropic IPO email offering me shares real? No email offering to sell you ANT shares directly is real. Anthropic shares trade on the Nasdaq through a broker. Any message asking you to pay, reserve, or verify a wallet to claim shares is a scam.
Need a disposable temporary email?
Protect your real inbox from spam and promotional clutter instantly and free.
Received a suspicious email, text, or charge alert?
Verify whether the sender domain is disposable, spoofed, or unverified before clicking links.
