Tutorial — Breach data analysis (defensive framing)
Last updated: October 21, 2026
Skill: breach-data-analysis-specialist
Time: ~30 minutes
Goal
Use breach metadata to assess account takeover risk or corroborate a public email — not to weaponize private data.
Steps
- Define question: “Was this account in breach X?” vs “What password?” — reject the second for Patriot use.
- Use Have I Been Pwned-style APIs or org-approved tooling.
- Record breach name, date, and classes of data disclosed (e.g. hashes vs plaintext — often unknown at summary level).
- Recommend rotation and MFA in defensive reports.
- Redact all secrets in notes.
- If research touches non-public figures, obtain legal review before retention.
What a breach record contains
Have I Been Pwned (HIBP), the service step 2 refers to, describes every breach with the same fields, and those fields are what a defensive report should cite. Its API returns the date of the breach, the date HIBP added it, the number of accounts affected (PwnCount) and a DataClasses list naming what was exposed — for the 2013 Adobe breach, email addresses, password hints, passwords and usernames. That list answers step 3’s question about what was disclosed without anyone handling the data itself.
Each record also carries flags that decide how much weight it can bear: IsVerified, IsFabricated, IsSensitive, IsRetired, IsSpamList, IsMalware and IsStealerLog. HIBP keeps unverified breaches in the system because, whatever their legitimacy, they still contain personal information about real people; it marks a breach fabricated when the data is known to be at least partly invented. Report the flags next to the breach name, and never present an unverified or fabricated breach as an established compromise.
HIBP’s own design follows the defensive framing. Breaches where someone’s presence could harm them if others found out — it lists 89, including Adult FriendFinder — are classed sensitive and can be searched only by the verified owner of the address, after proving they can receive email there. Retired breaches are not returned by an open search either. Password checks go through the separate Pwned Passwords service, which stores every password SHA-1 hashed and never next to an email address.
Pitfalls
- Paste dumps into chat tools — forbidden in most org policies.
- False negatives if email variants exist.
Next steps
Sources
- Have I Been Pwned, API documentation (v3) — breach model fields and flags.
- Have I Been Pwned, Frequently Asked Questions — sensitive, unverified and fabricated breaches; Pwned Passwords.
Expanded with AI research and drafting assistance from the primary sources cited; editorial review pending.
