Evidence: well sourced. Source-led synthesis; classifications are GeneticSecurity.org analysis, not findings by the source.
Case at a glance
- Case number
- 007
- Date / range
- Research capability
- Sector
- Forensic science
- Genetic asset
- Genetic markers used for trait inference
- Security principle
- Genetic Inference
Event summary
Genetic data can support probabilistic inference about externally visible traits and ancestry-related characteristics even when no direct identity match exists.
Acquisition and processing
Probabilistic prediction of selected visible traits; not direct identification
What became inferable or exposed
Genetic markers used for trait inference
Genetic data can support probabilistic inference about externally visible traits and ancestry-related characteristics even when no direct identity match exists.
Affected parties and consent
- Direct parties
- The unknown specimen source
- Indirect parties
- No indirect party count established for the bounded example
- Direct count
- Unknown / not assigned
- Indirect count
- Unknown / not assigned
- Consent status
- Authorization depends on collection and investigative context
Security dimensions
Confidentiality
Scope assessed from the specific asset and access described; no broader raw-genome disclosure inferred.
Integrity
No demonstrated data alteration established by the cited material.
Availability
No outage or destruction established in this case.
Provenance
Complete custody and processing evidence is unavailable; no numeric GPI rating assigned.
Read a prediction as a prediction
A trait estimate and an identification claim serve different purposes. A prediction may describe a characteristic that many people share. Even a well-supported estimate does not, on its own, name the source of a sample. Adding several estimates into a composite image does not remove the uncertainty attached to each component.
The visual form matters. People can read a face as a specific person even when the underlying analysis supports only a limited set of probabilistic traits. A responsible presentation should explain which features follow from the analysis and which reflect illustration, modeling choices, or assumptions.
Ask what was actually evaluated
When reading a claim about accuracy, look for the trait being predicted, the data used to evaluate it, and the conditions under which the result was measured. A strong result for one trait does not establish equivalent performance for another. Nor should a general performance statement be taken as certainty about an individual specimen.
These are questions for evaluating an inference, not an assertion that the cited research supplies one answer for every population or method. Where the case does not provide the relevant evaluation details, it should not manufacture a percentage to make the account appear complete.
Keep the scope of the example small
The GER-0 classification describes the bounded example of trait inference about one unknown source. It does not claim that phenotyping can never support broader inferences. If an assessment includes family matching, population characterization, or linked identity records, those additional activities change the scope and need their own rationale.
The practical lesson is to describe the output before scoring it. “A trait prediction from selected markers” tells a reader more than “a face from DNA,” and leaves room to discuss both the usefulness and the limits of the method.
GeneticSecurity.org analysis
Genetic Exposure Radius
Analysis: No indirect party count established for the bounded example. The rating describes possible scope, not harm or a count of affected people.
Confidence: moderate. Classification: GeneticSecurity.org analysis.
Genetic Persistence Risk
The described profile can remain useful for identification or comparison; retention and future linkability remain uncertain.
Confidence: low. Classification: GeneticSecurity.org analysis.
Genetic Provenance Integrity
Not assessed: the public sources do not establish enough of the complete biological and digital chain to score it. Unknown is not proof of missing controls.
Confidence: not assigned. Classification: GeneticSecurity.org analysis.
What this case does not prove
This does not establish that DNA phenotyping can reconstruct a person's exact face with photographic accuracy.
Mitigations and lessons
Report uncertainty and population limitations; validate each trait model; avoid presenting composites as photographs or identifications.
Primary sources
- PRIMARY SOURCE NIJ: What can DNA phenotyping tell us?
Secondary sources
No additional source listed. See the evidence notes for limitations.
Policy and standards
Genetic Security Policy and StandardsReview and correction history
Last reviewed: September 19, 2026. Initial source synthesis; no substantive corrections recorded.
Correction policy and logCite this case
GS-CASE-007. Forensic DNA Phenotyping: When DNA Becomes a Description. GeneticSecurity.org. https://geneticsecurity.org/cases/007-forensic-dna-phenotyping/