BYTETOOLS

CSV Anonymizer

Anonymise a CSV column by column: redact, partial mask, consistent SHA-256 pseudonyms, email masking, numeric jitter or date shifting — all in your browser.

Drop a .csv file here or click to browse

What is the CSV Anonymizer?

The ByteTools CSV Anonymizer removes personal data from a spreadsheet export so it is safe to share with a contractor, paste into a bug report or load into a test environment.

  • Per-column choice of redact, mask, pseudonym, jitter, date shift or drop
  • Consistent SHA-256 pseudonyms keep repeated values linked
  • Email masking that preserves the domain
  • Deterministic numeric jitter and fixed-offset date shifting
  • Random per-session salt you can regenerate at any time
  • Runs 100% locally — the raw file never leaves your device

How to use the CSV Anonymizer

  1. 1

    Paste your CSV or drop a .csv file, then confirm the delimiter and header row.

  2. 2

    Set a strategy for each column from its dropdown.

  3. 3

    Adjust the shared settings — redaction text, characters to keep, token length, jitter percent, date shift.

  4. 4

    Click Anonymize to run the transformation with Web Crypto.

  5. 5

    Copy the anonymised CSV or download it as a file.

About the CSV Anonymizer

The ByteTools CSV Anonymizer removes personal data from a spreadsheet export so it is safe to share with a contractor, paste into a bug report or load into a test environment. Paste your CSV and each column gets its own dropdown: keep it, redact it, mask part of it, pseudonymise it, jitter it, shift its dates, or drop it entirely.

The pseudonym option hashes each value with SHA-256 and a random per-session salt via Web Crypto, then truncates it to a short readable token. Because the same value always produces the same token, rows stay linkable and joins keep working — but the original value cannot be recovered from the output.

Every transformation happens in your browser and the raw file is never uploaded, which is the whole point: data that must be anonymised should not be sent to a server to be anonymised. Note that masking reduces risk but does not guarantee anonymity — a small dataset with rare combinations of values can still be re-identifiable.

Frequently asked questions

Will the same name always get the same pseudonym?

Yes, within one session. Each value is hashed with SHA-256 plus a random salt, so identical inputs produce identical tokens and your joins still work. Click 'New pseudonym salt' to break that link and start again.

Can the original values be recovered from the output?

Not from the token itself — SHA-256 is one way. But if an attacker can guess the salt and the value space is small, such as a list of known email addresses, a brute-force match is possible. Treat pseudonyms as a strong obfuscation rather than encryption.

Does anonymising a CSV make it GDPR-compliant?

Not automatically. Masking direct identifiers reduces risk, but a dataset can still be re-identifiable through rare combinations of remaining fields. Review the output against your own policy before sharing it, and drop columns you do not actually need.

What does numeric jitter do to my totals?

Each numeric value is nudged by up to the percentage you set, deterministically based on its hash, so the shape of the distribution survives while exact figures do not. Totals will not match the original — use it for realistic test data, not for reporting.

Is my file uploaded to be processed?

No. Parsing, hashing and writing all happen in your browser with the Web Crypto API. Nothing is transmitted, which is exactly why this is safe to use on data you are not allowed to send anywhere.

Related tools