In threat intelligence, a name can be assigned to a threat actor without specific guidelines. This leads to issues such
as a:
- A proliferation of threat actor names generating overlaps or different names for similar threat actors (e.g. some threat actors have more than 10 synonyms)
- Ambiguity in the words used to name the threat actor in different contexts (e.g. using common words)
- No clearly defined text format to describe the same threat actor (e.g. Is the threat actor name case sensitive? Is there a dash or a space between the two words?)
- Confusion between techniques/tools used by a threat actor versus its name (e.g. naming a threat actor after a specific malware used)
- Lack of source and list from vendors to describe their threat actor names and the reasoning behind the naming (e.g. did they name the threat actor after a specific set of campaigns? or specific set of targets?)
- Lack of time-based information about the threat actor name, such as date of naming or and UUID.
- Lack of open mirrored "registry" of reference, accessible to all, where to register a new threat actor name, or to access all already named threat actors. The "registry" can contain the time-based information mentionned above, it is a tool.
Before creating a new threat actor name, you **MUST** consider a review of existing threat actor names from databases such as the threat actor
MISP galaxy [@!MISP-G]. Proliferation of threat actor names is a significant challenge for the day-to-day analyst work. If your threat actor defined an existing threat actor, you **MUST**
reuse an existing threat actor name. If there is no specific threat actor name, you **SHALL** create a new threat actor following the best practices defined in this document.
When choosing a threat actor name, uniqueness is a critical property. The threat actor name **MUST** be unique and not existing in different contexts. The name **MUST** not be a word from a dictionary which can be used in other contexts.
The name of the threat actor **SHALL** be composed of a single word. If there is multiple part like a decimal value such as a counter, the values **MUST** be separated with a dash. Single words are preferred to ease the search of keywords by analysts in public sources.
The name of the threat actor **MUST** be expressed in ASCII 7-bit. Assigning a localized name to a threat actor **MAY** create a set of ambiguity about different localized version of the same threat actor.
The name of the threat actor **MUST NOT** be assigned based on the tools, techniques or patterns used by the threat actor. A notorious example in the threat intelligence community is Turla which can name a threat actor but also a malware used by this group or other groups.
Some known examples are included below and serve as reference for good practices in naming threat actors. The below threat actor names can be considered good example:
