Search filters this STIG's checks locally. Separate terms must all match (use quotes for an exact phrase). A search you ran on the Search page carries over here.
Assessment Title
Rule Text
Classified videoconferencing systems must authenticate with a unique user logon prior to performing functions and services.
Description
Discussion Text
DoD policy requires users to authenticate prior to being authorized to use available services. While requiring a user to authenticate to the video endpoint before it can be used to make or receive calls may detract from the video endpoint's "ease of use" and the "user experience" the capability should exist, be used where needed, and be configurable. Users should authenticate to activate the video endpoint for general use, make a call, or answer a call. Minimally, authentication should be a password unique to the user and recorded in session logs. Preferably, the video endpoint should support the use of DoD PKI for user authentication. To comply with DoD access control requirements for both users and administrators, a video endpoint should use a remote authentication server that can provide centralized management of passwords and accounts. This controls access to the videoconferencing system and limits the user’s privileges or authorizations. Many videoconferencing endpoints today do not provide sufficient identification, authorization, or auditing capabilities regarding their activation and use. While at least one vendor’s system can be configured to require the entry of a PIN to place a call, the feature is only a call accounting feature and not a security feature. While gatekeepers and gateways provide some access control, this control only relates to access to their services. They do not play a part in endpoint activation or use of the endpoint for point-to-point calls.
The ITU developed H.235 as the security recommendation for H.323 and other H.245-based systems. H.323 provides for user identification rather than device identification, using simple passwords/PINs or DoD PKI. H.235 has the capability of negotiating encryption and key exchange. The use of H.350 can improve security by providing standardized management and storage of authentication credentials, as well as multilevel authorization. The use of H.245 and H.350 in combination could be the solution to the endpoint activation and user identification deficiency currently exhibited by videoconferencing endpoints.
While it seems debatable whether a videoconferencing endpoint is, or should be, subject to DoD access control and auditing policies, particularly in unclassified environments, there are use cases where such compliance would be beneficial to the protection of DoD information. This is particularly in cases where a video endpoint is located in an area where classified materials, information, or discussions occur because an active video endpoint could generate a security incident. This issue could be more of a concern if the video endpoint was located in a classified work area while connected to an unclassified network or network having a lower classification than the work area. Compliance would also be beneficial for video endpoints in areas processing sensitive information. To protect the information, the video endpoint should remain dormant (even while powered on) and not capable of placing or answering a call unless it is activated by a user logging onto the system.
The ITU developed H.235 as the security recommendation for H.323 and other H.245-based systems. H.323 provides for user identification rather than device identification, using simple passwords/PINs or DoD PKI. H.235 has the capability of negotiating encryption and key exchange. The use of H.350 can improve security by providing standardized management and storage of authentication credentials, as well as multilevel authorization. The use of H.245 and H.350 in combination could be the solution to the endpoint activation and user identification deficiency currently exhibited by videoconferencing endpoints.
While it seems debatable whether a videoconferencing endpoint is, or should be, subject to DoD access control and auditing policies, particularly in unclassified environments, there are use cases where such compliance would be beneficial to the protection of DoD information. This is particularly in cases where a video endpoint is located in an area where classified materials, information, or discussions occur because an active video endpoint could generate a security incident. This issue could be more of a concern if the video endpoint was located in a classified work area while connected to an unclassified network or network having a lower classification than the work area. Compliance would also be beneficial for video endpoints in areas processing sensitive information. To protect the information, the video endpoint should remain dormant (even while powered on) and not capable of placing or answering a call unless it is activated by a user logging onto the system.
New to STIG checklists? Read What is a STIG checklist? and CKL vs CKLB.