Takeover without prior intrusion
One type of account takeover does not rely on Trojans, vulnerabilities, or phishing attachments. Attackers can create websites that appear legitimate and have compliant registration processes, collecting users' email addresses and passwords during voluntary registrations.

Summary
One type of account takeover does not rely on Trojans, vulnerabilities or phishing attachments. An attacker can build a website that looks legitimate, with a compliant sign-up flow, collect email addresses and passwords as users voluntarily register, then take them to other platforms for credential stuffing. Once the password of a main email account has been reused, the attacker can take over the inbox first, then intercept verification codes and reset the linked accounts. On the figures in section 3, roughly 1 in 2 adults (43% to 58%) are exposed to this, and more than 4 in 5 people aged 12 to 18.
For individuals, the key is not spotting every fake website, but this: never reuse your main email password; don't take new sites at face value by default; generate unique passwords with a password manager; and move two-factor authentication from SMS to an app or a hardware key.
1. Locating the problem
A site can contain no malicious code at all, and still have captchas, password-length rules and a full user agreement. The attack happens where business records meet user habits. HTTPS only shows that the connection is encrypted; it does not show that the operator won't record what you submit at the application layer. A "legitimate-looking interface" only shows that they are imitating a trustworthy product; it does not prove their motive.
2. A typical scenario
The attacker launches a site that looks like a membership site, a creator community or a download portal. The sign-up page often says, somewhere prominent:
We do not store plain-text passwords and do not offer recovery. Please remember yours.
The wording is close to what many real products do, but its effect is pressure: since the password can't be recovered, users are more willing to enter the one they know best, which is usually their main email password. After sign-up, the credentials are tried on many platforms. If the first thing to fall is the main inbox, the other accounts often don't need their passwords guessed: recovery and rebinding all lead back to that inbox.
3. Why it works
Password reuse is the default behaviour. The attacker doesn't need all your passwords, only the one that is reused.
• The main inbox is the root of your digital identity. Changing passwords, rebinding accounts, turning off 2FA: the usual recovery paths all point to it.
• A compliant appearance lowers your guard. Captchas and privacy policies are basic product hygiene, not proof of motive.
• "Cannot be recovered" punishes unique random passwords. The copy exploits a reasonable fear to push people back to "a password I know".
How common is this? In New Zealand, 43% of people use the same passwords for their online accounts (CERT NZ Cyber Security Behaviour Tracker 2024). There is no New Zealand figure for passwords that are only slightly changed, such as a capital letter swapped or a number added at the end, so this one is world-wide: in a study of 28.8 million users, 38% reused exactly the same password across sites and another 20% modified an existing one. A small change doesn't save you either: within 10 guesses the researchers could crack 30% of the modified passwords and every reused one (Wang et al., 2017).
So how many people are at risk? Being exposed only takes two things: a reused password, and an inbox that password can open. Two-factor authentication doesn't take you out of that group if the code, or the fallback for your key, is sent to that same inbox (see tip 3 below). On the figures above, that is 43% of New Zealanders counting exact reuse alone, and about 58% once slightly changed passwords are counted: roughly 1 in 2 adults. The only people who reuse passwords and still escape are those whose main inbox is itself protected by an authenticator app or a key, with no email or SMS fallback. No survey counts them, so the true share is somewhat below this.
Teenagers are more exposed still. There is no New Zealand figure for 12- to 18-year-olds, so this one comes from the largest survey of children's passwords so far, 1,505 US students in grades 3 to 12 (Theofanos, Choong and Murphy, NIST, 2021). 81% of middle-school students (grades 6 to 8, about ages 11 to 14) and 87% of high-school students (grades 9 to 12, about ages 14 to 18) said they use the same password for everything, and 39% and 45% said they share passwords with friends. By the same reasoning as above, that puts more than 4 in 5 people aged 12 to 18 at risk, and every friend who knows the password is one more place it can leak from.
4. Guidance for users (in priority order)
1. Use your main email password only for your main email
The inbox that receives verification codes for banking, payments, cloud storage and work must have a unique, strong password. If you can't be sure whether you've used it elsewhere, treat it as exposed: change it immediately on a trusted device, and check recent sign-ins and forwarding rules.
2. A unique password for every site, kept in a password manager
When you sign up to a new site, let the manager generate a long random password. When a site claims there is no recovery, that is all the more reason to store it in the manager rather than fall back to your main password. No recovery only means you have to keep it yourself; it doesn't mean you have to reuse one.
3. Move two-factor authentication off SMS where possible
In order of preference: passkey / hardware key → authenticator app (TOTP) → SMS. Download your recovery codes and keep them offline, somewhere other than your inbox. Don't assume that turning on verification codes means you can reuse your main password on untrusted sites. On accounts where you use a key, remove email and SMS as fallbacks where the service lets you, and put a key or an app on the email account itself.
In New Zealand, 43% of people always turn on two-factor authentication for their main accounts (NCSC survey of 1,011 adults, November 2025). There is no New Zealand breakdown by method; in a survey of 1,039 adults in the US and UK, 74% of people who had used two-factor authentication had used a code sent by email, and 85% a code sent by SMS (Duo, State of the Auth 2021). A code sent to your email is only as safe as that inbox: if the attacker already has your main inbox, the second factor goes straight to them.
A second factor is only as strong as the weakest way back into the account. Many accounts that accept a passkey or a hardware key still keep a fallback, such as a "try another way" option or a recovery link, that sends a code to your email. If that inbox's password is one you reused somewhere else, and the inbox itself isn't protected by a key or an app, the attacker never has to beat your hardware key: they sign in to the email, ask the other account for the fallback code, and read it there. So the risk below counts everyone who has used email codes, including people who also have an app or a hardware key.
How many people is that? No survey asks the same people about both habits, so this is an estimate that treats the two as independent, and it depends on how strictly you count:
• Strictest: 43% always use 2FA on their main accounts (NCSC) × 74% who have used email codes = 32%, × 58% who reuse or only slightly change passwords (Wang et al.) = about 18%.
• Middle: 51% of New Zealanders use 2FA on any of their accounts (InternetNZ Internet Insights 2023) × 74% = 38%, × 65% who reuse passwords (Google / Harris Poll, 2019) = about 25%.
• Broadest: 79% have used 2FA (Duo) × 74% = 58%, × 84% who reuse passwords across more than one site (Bitwarden World Password Day survey, 2022) = about 49%.
Somewhere between 1 in 5 and 1 in 2 adults have a second factor that can end up in an inbox guarded by a reused password.
4. Sign up with a separate email
Keep your main inbox for banking, government services, payments and work. Use a secondary inbox or an alias for communities, downloads and tools you only use now and then.
5. Ask three questions before entering a password
Does this site really need an account? Will I treat it as a long-term account? Am I about to enter a password I also use somewhere else? A polished interface and a captcha still don't answer these three questions.
6. Three things every quarter
Sign out sessions you don't recognise; check your email forwarding, filters, delegated access and app-specific passwords; check through a reliable service whether your email address has appeared in a breach (never paste your current password into a website you don't know).
7. Add recovery locks to high-value accounts
Payments, cloud drives, social media, domain and hosting consoles: turn on two-factor authentication that isn't SMS, and add a passkey first wherever you can. Preventing reuse matters more than being notified afterwards.
5. Conclusion
This threat works because three things are true at once: you can't verify whether an unfamiliar site records passwords; people enter their main password because "it can't be recovered"; and once the main inbox falls, the reset flows on every other site fall with it.
So the effective defences are just as concentrated: a unique password for the main inbox, random passwords everywhere else, a separate email for sign-ups, and stronger two-factor authentication; and if something is exposed, take back the inbox first, then change the other accounts. The goal is not to judge whether each website is acting in bad faith, but to make "even if this site is recording my password, it can't open the rest of my accounts" the default.

