Political Opinion & Affiliation

Detects references to political opinions and party affiliation: party membership, voter registration by party, political donations and campaign contributions. GDPR Article 9 lists political opinions as a special category of personal data with no fixed format — it is detectable only by topic vocabulary. Topic-grade detector, low confidence by design.

Type: regex
Engine: universal
Confidence: low
Confidence justification: Low by design: the attribute has no fixed format. Enforce tiers now require the attribute to be tied to an identified person record (a populated name/DOB/ID field), and 85 also requires a populated political field plus independent record evidence, so policy text, surveys-in-design and news naming the category stay at 65 discovery.
Jurisdictions: global, eu
Regulations: GDPR, Privacy Act 1988 (Cth)
Frameworks: ISO 27001, NIST CSF, SOC 2
Data categories: pii
Scope: broad
Risk rating: 7
Platform compatibility: Purview: Compatible, GCP DLP: Compatible, Macie: Compatible, Zscaler: Compatible, Palo Alto: Degraded, Netskope: Unsupported

Pattern

(?i)\b(?:political\s+affiliations?|political\s+opinions?|party\s+membership|party\s+affiliations?|voted\s+for|registered\s+(?:Democrat|Republican|Green|Libertarian|to\s+vote\s+as)|political\s+donations?|campaign\s+contributions?|political\s+party\s+member|partisan\s+affiliations?|voter\s+registration\s+party)\b

Corroborative evidence keywords

politics, election, candidate, ballot, donor, constituent, voter, electorate

Proximity: 300 characters

Should match

Campaign database - donor record Name: Avery Teston Member number: 0044718 Party code: GRN Donation amount: AUD 250 Contribution date: 12/07/2026 Note: political donation receipted; add to doorknock list. — Person record with populated political fields and donor-record evidence (85)
Constituent contact Name: Casey Rourke Customer ID: C4471820 Mentioned party membership during the call. — Person record naming party membership without a populated political field (75)
Electorate office CRM note - constituent follow-up Resident name: Jamie Okonkwo-Hart Voter ID: V-208815 Electorate: Hollis Ridge Voting intention: Undecided, leaning Independent Said she voted for the Greens at the last state poll; supporter score 3. Do not add to membership file. — Constituent CRM note with person fields, electorate and voting intention plus record evidence (85)

Should not match

The quarterly sales report shows revenue grew across all regions. — Unrelated business prose, no political terms
The committee registered the new vendor in the procurement system. — Near-miss prose ("registered" but not a party), no political opinion
The team selected the winning design after a quick internal poll. — Unrelated prose, "poll" without political affiliation vocabulary
Published election result includes aggregate party-support statistics. — Public aggregate political material is not a person-linked opinion record
Sample template: campaign donor record, amount AUD 100, contribution date 1 July. — Adjacent template language is not a populated political-opinion record
Membership file records appointment to the company board. — Sibling organisational membership does not satisfy the political primary
Policy guidance: this procedure explains how political opinion and party membership details are collected, stored and shared, who may access them and how long they are kept. Staff should read the privacy section before completing any form. It contains no details about any individual. — Policy/guidance text naming the topic without any populated record (65 discovery at most)
Membership application Name: ____________________ Member number: ____________________ Party: ____________________ Branch: ____________________ Signature: ____________ Date: __/__/____ — Blank form with unpopulated labelled fields (placeholder guard keeps it below 75)
News: the report on campaign contributions found that political donations from individuals rose sharply ahead of the state election. — News/training prose about the topic without a populated record (65 discovery at most)

Known false positives

Phrases like "voted for" or "registered" appear in non-political contexts (board elections, award votes, vendor registration), producing matches with no political-opinion content. Mitigation: Require corroborative keywords (election, candidate, donor, constituent, ballot) within the proximity window; treat isolated "voted for"/"registered" hits as low confidence.