Data sources
Every record is written by hand from a public source, and every record links to that source.
What counts as a source
We prefer sources in this order:
- Regulators and courts. Decisions, fines and court filings from data protection authorities, such as the Dutch Autoriteit Persoonsgegevens, the UK Information Commissioner’s Office, Ireland’s Data Protection Commission, the US Federal Trade Commission or the Office of the Australian Information Commissioner.
- The company itself. Official announcements, notices to customers and filings with financial regulators.
- Established news organisations, only when neither of the above is available, and the record is then marked as an unverified report unless the company or a regulator has confirmed it.
We don’t use leaked data, hacker forums or anonymous claims as sources, and we never publish the leaked data itself.
Confirmed or unverified
| Label | What it means |
|---|---|
| Confirmed | The company or a regulator has said the incident happened. |
| Unverified report | The incident has been reported, but not confirmed by the company or a regulator. |
What a record contains
- When the breach happened (or started), as precisely as the source allows
- How many people were affected, using the figure in the source
- The types of personal information exposed
- A short summary in plain language
- A link to the source
Where a court case is still going on, the record says so, and describes what a regulator alleges as an allegation, not a finding.
How risky is the exposed data?
Each type of exposed information is coloured by how much harm it can do to the people affected if criminals get hold of it:
| Level | Examples | Why |
|---|---|---|
| ● High risk | Passwords, payment card details, bank account numbers, passport, ID and social security numbers, health and genetic information | Can be used directly for fraud or identity theft, or is very sensitive and can’t be changed. |
| ● Medium risk | Dates of birth, home addresses, phone numbers, locations, booking details | Makes scams, phishing and identity theft more convincing, especially combined with other data. |
| ● Lower risk | Names, email addresses, user IDs, loyalty numbers | Often already public, but still useful to spammers and scammers. |
Each breach is labelled by the most serious type of data it exposed. These levels are a general guide, not a measure of the risk to you personally.
Track record
Each company has a short track record at the top of its page. It sums up the records below it, and is not a security rating:
- Breaches on record: how many breaches we have recorded, and in which years.
- Time to make it known: how long it took between the breach starting and the company first making it known, to the public or a regulator. Often the delay is because the company did not notice the breach, not because it kept quiet. Where nobody knows when a breach really happened, we don’t show a time.
- Regulator or court action: fines, settlements, court cases, orders and investigations by regulators or courts, as reported by official sources. “None recorded” means we haven’t found any, not that none exist.
- Most sensitive data exposed: the highest risk level of any data exposed in the company’s breaches.
- Official statements: how many of the breaches the company published its own statement about.
Official company statements
Where the company published its own statement about a breach, the record links to it, so you can read what the company itself said. If we couldn’t find one, the record says so.
Corrections
If something in a record looks wrong or out of date, please tell us with a link to a public source. Companies named in a record are welcome to send corrections too. We aim to review every request and fix confirmed mistakes promptly.