NordPass 2026: turn password rankings into access controls
A ranking of exposed credentials can help teams choose useful checks. The enterprise response needs to cover each account’s entry and recovery paths.
Lire en portugaisBy Pentesty
Understand what the research measured
The official NordPass report identifies its eighth edition and places 123456, admin and 12345678 in the top three positions. The company says 123456 has led seven of the eight editions. Reports published on November 17 by Cyber Security Brazil and CiberSegurança.pt concern that same research, rather than two separate incidents.
The current methodology describes passwords exposed between September 2025 and September 2026 in public breaches and dark web repositories, organized across 44 countries. The work brings together NordPass, NordLayer Intelligence by NordStellar and independent researchers. It is a sample of exposed credentials, not a user census; a ranking does not establish how many accounts remain compromised today.
The dynamic count table was unavailable when we consulted the primary page. We therefore reproduce no totals or country rankings. The page identifies its 2026 edition and textual findings, but we could not confirm an exact release date there. We also do not present the research as proof that AI caused new intrusions.
Turn the list into an acceptance condition
We recommend mapping where the organization still accepts passwords: its identity provider, custom applications, administrative interfaces and older systems. Choose test accounts and define which inputs should be rejected. Do not ask employees for their real passwords to compare them with the ranking.
- Check whether registration and password changes reject common passwords or those known to be compromised. Preserve the applied policy and the result from a controlled account; a written rule does not establish enforcement on the actual path.
- Test controls that limit attempts and signal suspicious patterns within an authorized window, without affecting production accounts. Record how the team distinguishes an isolated rejection from a sequence that requires investigation.
- Confirm that services support unique, long passwords and password managers. NordPass recommends at least fifteen characters; technical limitations need documented exceptions, with an owner and a remediation plan.
Cover recovery, second factors and devices
Enabling MFA or offering passkeys is part of the work. Check enrollment of a new factor, loss of a device and account recovery: who approves, what evidence is required and when a previous session stops working. Use controlled identities and preserve those decision records without collecting real people’s codes or secrets.
The same review should identify integrations and service accounts that do not follow the interactive login. A password change can leave other access methods active. Our case on cloud vendor credentials helps organize that inventory by identity and resource.
Consider software operating on the workstation as well. Sign2Shell research into signing components illustrates why an authentication policy does not independently address the behavior of extensions and native programs. These are complementary checks with their own evidence.
Track an exception until it is retired
Set a periodic review for exceptions and repeat the checks when the provider, application or recovery workflow changes. A useful record links the rule to the tested path, date, owner and result. The ranking offers a starting signal; completing the work depends on establishing what the environment accepts and addressing the gaps found.
Protect the resulting evidence as well: record test-account identifiers and outcomes without attaching real passwords. Agree who may consult these records and how long they will be needed to verify the correction in the assessed environment.
Frequently asked questions
Does the list describe every user’s password?
No. It draws on breaches and repositories examined by the researchers. The sample does not automatically represent the entire population.
Is this a new edition or the 2025 list republished?
The official page identifies the eighth, 2026 edition and a research window from September 2025 to September 2026.
Does this analysis confirm victim counts in Brazil?
No. We could not verify the dynamic count table or country rankings. A password ranking is not a count of current victims either.
Does changing a password end every possible access path?
Do not assume so. Check sessions, recovery, additional factors and integrations within the account’s scope, retaining evidence for each check.
Sources and references
- Cyber Security Brazil: editorial news reference
- CiberSegurança.pt: 2026 edition and limits of the available data
- NordPass: 2026 edition, methodology and recommendations
Original Pentesty analysis, published November 17, 2026. The proposed scenarios are editorial guidance, with scope and validation to be defined for each environment.