Derivative classification hinges on how markings flow from the source to anything you create from it. The derivative material must reflect the source's classification markings. Those markings communicate how sensitive the information is and how it must be protected, shared, and handled. If you derive material from a source that carries specific classification and dissemination marks, those cues should appear on the derivative to preserve its provenance and proper protection. This helps ensure that anyone who handles the derivative understands its sensitivity level and complies with the original constraints. The other ideas don’t fit because they imply you can freely reinterpret, omit, or ignore the source's markings, which would misrepresent the information's sensitivity and violate protection requirements.

Declassification directives provide rules for releasing information, including derivatives, but they don’t set the derivative’s current sensitivity level. The level of a derivative is determined by applying the present classification guidance to the transformed material—usually reflecting the highest level of any source material and how the derivative was created. A declassification directive can indicate when or under what conditions the material should be released in the future, but it does not automatically assign or change the derivative’s level today. So these directives are useful for guiding declassification decisions, while the derivative’s level remains governed by current guidance.

When a derivative is created from sources at different classification levels, the overall classification is driven by the most restrictive source. That means the derivative must be reclassified to the highest level represented and access restricted to those who have that level of clearance (and the necessary need-to-know). This protects the entire derivative from being disclosed more than allowed by the most sensitive source it contains. Distributing at the lower level would create a risk of exposing information at a higher level than allowed, which is why that approach is not appropriate. Creating separate copies for each level would complicate safeguarding and could lead to misallocation or leakage. Sharing only the portion common to all sources would omit legitimate, potentially sensitive information unique to higher-level sources, undermining the integrity of the derivative. The correct approach keeps the derivative as the highest level and controls access accordingly.

Security Classification Guides establish the exact criteria and levels you use when deriving classifications for different kinds of information. In derivative classification, you take information that’s already classified and determine the appropriate new classification for material that incorporates or summarizes it. SCGs provide the authoritative rules for each information type, specifying what content or contexts trigger particular classification levels and how those levels should be applied to derivatives. This guides you to apply consistent, appropriate protection across documents, helping prevent both over-classification and under-classification. SCGs act as a reference so others can reproduce the same decision in similar situations and know what markings or safeguards to apply. They are not mere formatting guidelines, they don’t restrict sharing within an agency by themselves, and they don’t decide who has need-to-know. The role of SCGs is to supply the criteria and levels used to derive classifications for different information types.