Compliance first: We never take HIPAA/PHI data or personal customer data. Everything is de-identified before it leaves your control.
Home / Guidelines / Open data policy guideline 07
Open data policy guideline 07

Sensitive Information: Open Data Policy Guideline 7

Sensitive Information: Open Data Policy Guideline 7

Guideline 7 asks open data policies to protect privacy and security with narrow, clearly defined exemptions, and to release sensitive information in aggregate where individual records cannot be published.

Balancing openness and protection

Openness and privacy are not opposites, but they do have to be balanced. A good open data policy inherits the exemptions already in public records law and adds safeguards for risks specific to bulk online publication, where data sets can be combined in ways that expose individuals.

The guideline makes three practical points. Individual-level data about private persons deserves special scrutiny. Aggregation often lets a government release useful information without exposing anyone. And exemptions should be drawn as narrowly as possible, with a balancing test rather than a blanket refusal.

How policies typically phrase it

  • A definition of "protected" or "restricted" data that cross-references existing law.
  • A list of excluded categories, such as security-sensitive infrastructure plans or personal identifying information.
  • A requirement that publication be reviewed for privacy before release.
  • A preference for releasing aggregated or redacted versions when the raw data cannot be published.

Places whose policies adopted it

Charlotte's 2015 policy spells out excluded categories in detail, including federally defined sensitive security information and building plans for city infrastructure. New York City's 2012 Local Law 11 excludes data that agencies may withhold under state public officers law. Los Angeles's 2013 directive publishes data to the extent permitted by law and subject to valid privacy, confidentiality and security restrictions.

The guideline is tagged in most policies in this index, including Chicago, Philadelphia and South Bend.

Common pitfalls

  • Overbroad exemptions that let departments withhold anything inconvenient.
  • Ignoring re-identification, where separately harmless data sets combine to reveal individuals.
  • No aggregation path, so sensitive data is simply withheld rather than summarized.

Practical safeguards before publication

Policies that handle this guideline well usually pair the legal exemptions with an operational review step. Common elements include:

  • A named reviewer, often legal or privacy staff, who signs off before a new data set is published.
  • A checklist covering direct identifiers, quasi-identifiers such as exact dates and addresses, and small-cell counts.
  • Standard techniques such as aggregation by area or time, rounding, generalizing locations to blocks, and suppressing rare categories.
  • A process for removing data quickly if a problem is found after release.

The checklist approach recognizes that privacy risk depends on context. A data set may be safe on its own but risky when joined to other public records, so reviewers consider what else is already available. Federal guidance on personally identifiable information, such as NIST SP 800-122, is a common reference point.

Why this matters for your business data

This is the guideline closest to how private data should be shared. Responsible data sales rest on the same principles: exclude what must be protected, aggregate or transform the rest, and check for re-identification risk before anything leaves the building.

Our guide to de-identified data explains the techniques, and open data policy to data value shows how public-sector safeguards translate to private operational data. We never take HIPAA/PHI data or personal customer data. Everything is de-identified before it leaves your control.

Our compliance promise: We never take HIPAA/PHI data or personal customer data. Everything is de-identified before it leaves your control.

Frequently asked questions

What counts as sensitive information in an open data policy?

Typically personal identifying information, data protected by privacy statutes, security-sensitive infrastructure details, law enforcement records exempt from disclosure and anything confidential under state or federal law.

How can a government release data that contains personal details?

Often through aggregation or redaction: publishing counts, rates or summaries by area or time period instead of individual records, after reviewing whether the result could identify anyone.

What is re-identification risk?

The risk that data stripped of names can still be linked back to individuals, for example by combining it with other public data sets. Responsible publishers check for this before release.

Should exemptions be listed in the policy?

Most strong policies either list them or cross-reference existing law, and the guideline recommends keeping them as narrow as the law allows.

Policies in the index that reflect guideline 7

PlaceAdoptedInstrumentGuidelines
Portland, ORSep 30, 2009Council resolution37810162230
San Francisco, CAOct 29, 2010Ordinance47811182526
Cook County, ILMay 4, 2011Ordinance12378131825
New York, NYMar 7, 2012Local law (Local Law 11 of 2012)14578101118
Philadelphia, PAApr 26, 2012Executive order (No. 1-12)34578182224
Madison, WIAug 15, 2012Ordinance457810111318
Chicago, ILDec 10, 2012Executive order34678101113
Montgomery County, MDDec 17, 2012County legislation (Bill 23-12, Open Data Act)457811131822
San Francisco, CAMar 28, 2013Ordinance34578101118
South Bend, INAug 22, 2013Executive order237810182528
Austin, TXAug 26, 2013Administrative memo123457810
Louisville, KYOct 11, 2013Executive order (No. 1, Series 2013)1234781018
Oakland, CAOct 15, 2013Council resolution (No. 84659)2345781011
Honolulu, HINov 27, 2013Ordinance237810111820
Los Angeles, CADec 18, 2013Executive directive (No. 3)37818222630
Pittsburgh, PAMar 19, 2014Ordinance12345678
Hartford, CTMar 26, 2014Executive order34789182224
Nashville, TNMay 12, 2014Executive order (No. 043)2345781011
Chattanooga, TNMay 31, 2014Executive order123467810
Phoenix, AZJun 18, 2014Administrative policy13726
Washington, DCJul 21, 2014Executive directive (Mayor's Order 2014-170)12356789
Houston, TXSep 18, 2014Administrative policy (AP 8-7)234567810
Charlotte, NCJan 1, 2015IT policy2781924
Denton, TXFeb 9, 2015Administrative directive357811121618
Boston, MAJul 30, 2015Administrative policy478132425
Suffolk County, NYJul 31, 2015Administrative policy78131820222426
Jackson, MSSep 1, 2015Executive order (No. 2015-06)12345678
Cambridge, MASep 21, 2015Ordinance12345678
Durham, NCNov 23, 2015Administrative policy7818192430
Seattle, WAFeb 1, 2016Administrative policy (CTO policy)14678101112
Seattle, WAFeb 26, 2016Executive order17242627
Baltimore, MDMar 7, 2016Ordinance1781113182426
Las Vegas, NVApr 12, 2016Administrative policy (City Manager policy CM510)12345678
Saint Paul, MNApr 27, 2016Council resolution12578101113
Milwaukee, WIJul 14, 2016Council resolution12356789
Bellevue, WAAug 1, 2016Management policy15678101118
Naperville, ILNov 15, 2016Council resolution45789101112
Fort Collins, COFeb 7, 2017Council resolution12345678
Knoxville, TNMay 2, 2017Executive order14578111319
Syracuse, NYJul 26, 2017Executive order1456781113

Sources

  1. Sunlight Foundation, Open Data Policy Guidelines (historical reference)
  2. NIST SP 800-122, Guide to Protecting the Confidentiality of PII
  3. Data.gov Privacy and Security Checklist (archived)
  4. Sunlight Foundation blog on balancing tests for sensitive information (2014)

Updated December 31, 2024.