Password Statistics 2026: Security, Reuse, Passkeys and MFA

Written by Branko Krstic • Edited by Neyza Albarracin
• 10 min read

If you’ve ever used passwords like “123456” or “123455678” to protect your online account, you’ve probably had a stolen password at some point.

It’s a shame to still rely on passwords like “iloveyou” in the age of IoT gadget abundance when your machine and your fridge are exchanging data faster than you can think.

Unfortunately, far too many people rely on bad practices, making it extremely easy for hackers to access data that does not belong to them.

Keep reading this list of staggering password statistics below to learn about other people’s mistakes and their consequences.

Staggering Password Statistics to Keep in Mind (Editor’s Pick)

  • NIST SP 800-63B requires passwords used as a single authentication factor to be at least 15 characters; passwords used only inside MFA may be as short as eight.
  • NIST also prohibits services from forcing periodic password changes, and from imposing composition rules such as a required mix of character types.
  • NordPass’ 2025 analysis of breached and dark-web data found “123456” was again the world’s most common password, topping its global list in six of seven editions.
  • Microsoft reports that more than 97% of the identity attacks it observed were password spray or brute force.
  • Verizon’s 2026 DBIR found 31% of breaches began with software-vulnerability exploitation, the first time it has overtaken credential abuse as the leading entry point.
  • The same report still finds credential abuse somewhere in the attack chain in 39% of breaches, so passwords have not stopped being the problem.
  • Bitwarden’s 2025 survey of 2,300+ employed adults in six countries found 72% of Gen Z respondents reused passwords, against 42% of Boomers.
  • The FIDO Alliance estimated 5 billion passkeys were in active use in May 2026.
  • In FIDO’s 10-country consumer survey, 90% of adults knew about passkeys, 75% had enabled one, and 49% used them regularly when offered.
  • In FIDO’s workforce survey, 68% of organisations with 500+ employees were deploying, piloting or rolling out passkeys for staff sign-in.

Password Statistics 2026

Two things changed the password conversation recently. The security standard that most organisations follow was rewritten, and passkeys stopped being a prediction and became something measurable.

Every figure below carries the year of the study or standard it comes from, because a 2026 page is not the same thing as 2026 data.

1. “123456” was again the world’s most common password in 2025.

(Source: NordPass, 2025 edition)

NordPass puts it bluntly: “123456” has “once again claimed the controversial title of the world’s most common password”, which makes six years out of seven that it has topped the list. “password” has taken the top spot just once.

Read the methodology before you read the ranking. NordPass and NordStellar analysed recent public data breaches and dark-web repositories from September 2024 to September 2025, organised by country of origin across 44 countries. That tells you which weak passwords turn up most often in stolen credential dumps. It does not tell you what share of all internet users chose each one, and passwords that never leak never appear.

2. NIST requires single-factor passwords to be at least 15 characters.

(Source: NIST SP 800-63B)

The wording in the current revision is a hard requirement, not a suggestion: verifiers and credential service providers “SHALL require passwords that are used as a single-factor authentication mechanism to be a minimum of 15 characters in length”. Where a password is only one part of multi-factor authentication, services may allow shorter ones but SHALL still require at least eight characters.

This is a standard that services must meet, not a description of what people currently do. The old version of this page cited an eight-character average and told readers to use “more than eight characters”; the floor for a password doing the job alone is now nearly double that. Length is also the single most useful thing you control, because every additional character multiplies the work a brute-force or credential-cracking attack has to do.

3. NIST says services should not force periodic password changes.

(Source: NIST SP 800-63B)

The standard reads: verifiers and CSPs “SHALL NOT require subscribers to change passwords periodically”, but “SHALL force a change if there is evidence that the authenticator has been compromised”.

That reverses advice this page itself used to give, and that plenty of workplaces still enforce. The reasoning is behavioural: people forced to invent a new password every 90 days produce predictable variations, so the organisation trades a small theoretical gain for passwords an attacker can guess from the last one. Change a password when there is reason to think it leaked, not because a calendar said so.

4. NIST says services should not impose character-composition rules.

(Source: NIST SP 800-63B)

The same section is explicit: verifiers and CSPs “SHALL NOT impose other composition rules (e.g., requiring mixtures of different character types) for passwords”.

It is worth being careful about who this applies to. It tells services to stop demanding one uppercase, one number and one symbol. It does not tell you that randomness is bad: CISA still recommends long, random, unique passwords for individuals, and a password manager generating random strings remains the best option available to you. The distinction is between what a service may require and what you may choose, and the two are easy to confuse.

How Password Attacks and Breaches Actually Start

The old version of this section mixed attack telemetry, breach reports, ransomware frequency and survey anxiety together as though they measured one thing. They do not, so the figures below say what population each one covers.

The short version: passwords are still overwhelmingly how attackers try to get in, but they are no longer the single most common way breaches actually begin.

5. More than 97% of the identity attacks Microsoft observed were password spray or brute force.

(Source: Microsoft Digital Defense Report 2025)

Microsoft’s telemetry shows the overwhelming majority of identity attacks are unsophisticated guessing at scale: password spray, where one common password is tried against many accounts, and brute force. The more advanced techniques that get the headlines, token theft, adversary-in-the-middle and attacks on MFA infrastructure, together account for under 3%.

Two cautions. This is a share of identity attacks Microsoft saw, not a share of all breaches, and attempts are not successes. It does say something useful though: the attack you are most likely to face is someone trying “123456” against thousands of accounts, which is precisely the attack that long, unique passwords defeat.

6. 31% of breaches began with vulnerability exploitation in the 2026 DBIR.

(Source: Verizon 2026 DBIR)

For the first time in the report’s history, exploiting a software vulnerability was the most common first known entry point into a breach, at 31%, up from 20% a year earlier. It overtook both phishing and credential abuse, which fell from 22% to 13%.

Part of that fall is measurement rather than behaviour: this edition added pretexting as a tracked entry vector, and because pretexting often involves credentials, the new category absorbed some of credential abuse’s share. Measured the way the previous edition counted, credential abuse would have come in at 16%. Our data breach statistics cover the rest of the report’s findings.

7. Credential abuse still shows up in 39% of breaches.

(Source: Verizon 2026 DBIR)

This is the figure that stops the previous stat being misread. Counting credential abuse anywhere in the attack chain rather than only as the first step, the same report finds it in 39% of breaches.

So “vulnerabilities overtook passwords” is true and narrow. It describes which door attackers opened first. Stolen credentials remain one of the most common things they use once inside, for moving between systems and escalating access, which is why none of this makes password hygiene less important. Anyone claiming that over 80% of breaches are caused by weak passwords, as this page once did, is also wrong in the other direction.

8. The 2013 Adobe breach exposed 153 million accounts.

(Source: UpGuard)

In October 2013, 153 million Adobe accounts were compromised. Internal IDs, usernames, emails, poorly encrypted passwords and, damagingly, plain-text password hints were all exposed.

This is a historical example, not current practice, and it is here because it illustrates something specific: the encryption failure was bad, but the password hints were worse, since they let attackers guess credentials that the encryption had not actually given up. Thirteen years on it remains one of the clearest demonstrations of why a password’s security depends on how the service stores it, something no amount of care on your side controls.

Password Security Statistics

Despite the fact that cybersecurity risks have evolved, businesses continue to depend on passwords alone as an efficient protection measure.

The password statistics below reveal just how risky this outdated thinking can be.

9. 72% of Gen Z respondents reused passwords, against 42% of Boomers.

(Source: Bitwarden World Password Day survey, 2025)

Bitwarden surveyed more than 2,300 employed adults across the United States, Australia, the United Kingdom, Germany, France and Japan in 2025. Password reuse ran highest among the youngest workers: 72% of Gen Z respondents, against 42% of Boomers.

Keep the scope visible. This is a vendor-sponsored survey of employed adults in six countries, so it describes those respondents rather than every young person worldwide, and self-reported behaviour tends to understate rather than overstate bad habits. The direction is still the interesting part, since it is the reverse of what “digital native” assumptions would predict.

10. 59% of Gen Z reuse a password even at a company that has been breached.

(Source: Bitwarden World Password Day survey, 2025)

From the same survey: 59% of Gen Z respondents said they reuse an existing password when updating an account with a company that has suffered a data breach, compared with 23% of Boomers.

This is the more alarming of the two findings. A breach notification is the one moment when the need for a fresh, unique password is unambiguous, and a majority of the youngest group reported recycling one anyway. It is also the exact behaviour that turns a single breach into credential stuffing against every other account.

11. Password choices barely differ across five generations.

(Source: NordPass, 2025 edition)

NordPass broke its 2025 breach data down by age group and found, in its words, that “the password habits of an 18-year-old are strikingly similar to those of an 80-year-old”. The most common choice in every single generation was either “12345” or “123456”.

  • Generation Z: 12345, 123456, 12345678
  • Millennials: 123456, 1234qwer, 123456789
  • Generation X: 123456, 123456789, 12345
  • Baby boomers: 123456, 123456789, 12345
  • Silent generation: 12345, 123456

Set this beside the Bitwarden findings above and there is a real tension worth naming. The surveys say younger people reuse passwords more; the breach data says everyone picks equally terrible ones. Both can hold, because they measure different failures, reuse versus choice, from different evidence, self-report versus leaked credentials.

12. 43% of US internet users had shared a password with someone else.

(Source: Google/Harris Poll, 2019)

A Harris Poll conducted for Google surveyed 3,419 US adults between August 16 and 29, 2019. It found 43% had shared a password with another person, 66% reused passwords across accounts, and 59% used a name or birthday in a password.

This is dated and included as historical sharing behaviour, not a 2026 rate. It is kept because password sharing is otherwise absent from current research, and because the follow-up detail has stayed memorable: most sharing was with a partner, yet only about one in ten changed those credentials after a relationship ended.

13. Phishing-resistant MFA can stop over 99% of these attacks.

(Source: Microsoft, CISA)

Microsoft’s 2025 reporting says phishing-resistant multi-factor authentication can stop over 99% of password-based attacks of the spray-and-brute-force kind described above, even when the attacker already has the correct username and password.

The wording matters, and the old claim on this page that MFA “blocks 99% of all password safety issues” was far too broad. It is not 99% of all cyberattacks, breaches or password problems. It is a specific, very large reduction against a specific and very common class of attack. Not all MFA is equal either: CISA recommends FIDO/WebAuthn as phishing-resistant authentication and treats passkeys as an acceptable FIDO option, while warning that SMS codes are not phishing resistant, because an attacker running a convincing fake login page can collect the code along with the password.

Previous versions of this page predicted that passwords would disappear within five years and that 2022 would be the year MFA became the norm. Both dates have passed, so the forecasts are gone and measured adoption has replaced them.

Passkeys replace a password with a cryptographic key pair held by your device, unlocked with a fingerprint, face or device PIN. Nothing reusable is typed, so there is nothing for a spray attack or a fake login page to capture.

14. The FIDO Alliance estimated 5 billion passkeys in active use.

(Source: FIDO Alliance, May 2026)

On World Passkey Day, May 7, 2026, the FIDO Alliance put active passkeys at 5 billion. Passkey support now ships across the major operating systems, browsers and password managers, which is what moved the number at this scale.

This is an Alliance estimate rather than an audited count, and the Alliance is the industry body promoting the standard, so treat the precise figure with the caution that deserves. Note too that a passkey in existence is not a password retired: most accounts that gained a passkey still accept the old password as a fallback.

15. 75% of consumers have enabled a passkey, but only 49% use them regularly.

(Source: FIDO Alliance consumer survey, April 2026)

FIDO surveyed 11,000 adults across ten countries (the US, UK, France, Germany, Australia, Singapore, Japan, South Korea, China and India) in April 2026, with a margin of error of ±0.9 points. It found 90% were familiar with passkeys, 75% had enabled one on at least some accounts, and 49% used them regularly when the option was available.

The gap between 75% and 49% is the real finding. Enabling a passkey is a one-off prompt most people accept; using it consistently means it has to be the easier path every time, and often it is not, because the password field is still sitting right there. Adoption is genuine and substantial, but passwords have not been eliminated, and NIST is explicit that passwords are not phishing resistant.

16. 68% of organisations are deploying, piloting or rolling out passkeys for staff.

(Source: FIDO Alliance workforce survey, April 2026)

FIDO’s companion workforce survey covered 1,400 authentication decision-makers at organisations with 500 or more employees across the same ten countries, with a margin of error of ±2.6 points. Of those, 68% said their organisation was deploying, piloting or rolling out passkeys for employee authentication.

Read that as 68% of the organisations these decision-makers represent, not 68% of all businesses: the sample is deliberately large employers, who move on authentication long before small ones do. “Deploying, piloting or rolling out” also covers everything from a finished migration to a trial with one team.

Protecting Your Accounts: What Current Guidance Says

Passwords are attacked in a few well-understood ways. Brute force tries combinations until one works. Password spray inverts that, trying a handful of very common passwords against a great many accounts. Credential stuffing replays username and password pairs leaked from one service against others. Hash cracking works offline on stolen password databases, which is why how a service stores your password matters as much as how you chose it.

The guidance below follows current NIST and CISA advice. It differs from what this page previously recommended, in particular on scheduled password changes, which NIST now tells services not to require.

  • Make them long. Length beats complexity. NIST’s floor for a password doing the job on its own is 15 characters, so treat that as a minimum rather than a target. A passphrase of several unrelated words is easy to remember and hard to crack.
  • Make every one unique. This is the single highest-value habit, because it is what stops one breach becoming many. Credential stuffing only works on reused passwords.
  • Use a password manager. Unique 15-character passwords for dozens of accounts are not memorable by design, which is the point: let software generate and store them so you only remember one strong passphrase.
  • Change a password when there is a reason to. Evidence of compromise, a breach notification, or a password you know you have reused. Not on a 90-day schedule, which pushes people into predictable variations of the same password.
  • Check whether your passwords have already leaked. Compromised-credential screening is built into most password managers and browsers. A password that is long, unique and already in a breach dump is not protecting anything.
  • Turn on MFA, and prefer the phishing-resistant kind. Any MFA is far better than none. CISA recommends FIDO/WebAuthn, including passkeys, for accounts that matter most, and cautions that SMS codes are not phishing resistant, since a convincing fake login page can harvest the code along with the password. An authenticator app sits between the two.
  • Use passkeys where they are offered. Nothing reusable is typed, so there is nothing for a spray attack or a fake login page to capture. Most accounts keep the password as a fallback, so this adds a safer route rather than closing the old one.
  • Check where you are typing it. Credential phishing depends on you not looking at the address bar. Confirm the domain is the one you meant, and that the connection uses a valid SSL certificate, before any password goes in.

For more tips, read our comprehensive guide on how to protect your online identity.

Wrap Up

But what is the most important lesson from these fascinating password stats?

Two of them, really. The guidance changed: length and uniqueness now matter, forced rotation and mandatory character mixtures do not, and the standard says so in as many words. And passkeys became measurable, with 5 billion of them in use and two thirds of large employers rolling them out, while still not displacing the password underneath.

What has not changed is the underlying behaviour. NordPass found “123456” on top for the sixth year in seven, and found an 18-year-old picking much the same password as an 80-year-old. The weak link was never the technology.

When it comes to accounts that have little to no confidential information, it may seem like a minor problem. But the password statistics above clearly show the danger you’re exposing yourself to.

If you use the same password for several accounts, cracking one of those less-important accounts unlocks the door to your digital life. If you haven’t developed strong password habits, take the first step now.

About Branko Krstic

Author

Branko is a round-the-clock tech geek and loving it. His ideal vacation destination is the Akihabara District (or really any place he can take his computer). If there’s a server out there, count on him to find out what it’s made of… and tell you all about it.

View all posts by Branko Krstic →

About Neyza Albarracin

Editor

Editorial contributor at WebTribunal.

View all posts by Neyza Albarracin →