PAN detection
A publicly auditable detection pattern in the LeakSnitch engine. Category: Personal Data (PII). Severity: High.
Short answer
LeakSnitch detects pan with a base score of 85 (High severity). The pattern below is fully public, so any security team can verify exactly what gets flagged before rollout.
\b([A-Z]{5}[0-9]{4}[A-Z])\bBase score
85 / 100
Severity
High
Category
Personal Data (PII)
Why it matters
Indian Permanent Account Numbers follow a strict five letter, four digit, one letter format that this pattern enforces exactly.
In the LeakSnitch engine this pattern is one of 58 public rules combined with context signals, entropy analysis, and checksum validators. A single match never blocks by itself; the full context aware scoring decides whether pan in a prompt is a real leak or a false positive.
Frequently asked questions
Why is the pan pattern public?
LeakSnitch's competitive advantage is the context aware engine, not hidden regexes. Publishing every pattern lets security teams audit, test, and trust the detection before it ships to their organization.
Can I test this pan pattern?
Yes. The pattern is public and auditable. You can copy the regex from this page, run it against sample text in any tool, and see the exact format match for yourself. LeakSnitch validates each match with context signals, entropy analysis, and checksum checks before blocking.
Does LeakSnitch use other signals for pan?
This public pattern is the format based layer. The engine adds context signals, disclosure phrasing, entropy fallback, and checksum validation to confirm high confidence before blocking.
Related patterns
Stop AI leaks on the device, not in the cloud.
Install LeakSnitch in under 30 seconds and protect every prompt sent to ChatGPT, Claude, Gemini, and 25+ AI tools. Free for individuals.