People's Newsroom

SHARING AGREEMENTS AND REPORTING REQUIREMENTS IN COMPUTER SECURITY DEFENSE

Organizations trying to share information with external organizations should consult with their legal department before initiating any coordination efforts. There may be contracts or other agreements that need to be put into place before discussions occur. An example is a nondisclosure agreement (NDA) to protect the confidentiality of the organization’s most sensitive information. Organizations should also consider any existing requirements for reporting, such as sharing incident information with an ISAC or reporting incidents to a higher-level CIRT.

Information Sharing Techniques. Information sharing is a key element of enabling coordination across organizations. Even the smallest organizations need to be able to share incident information with peers and partners in order to deal with many incidents effectively. Organizations should perform such information sharing throughout the incident response life cycle and not wait until an incident has been fully resolved before sharing details of it with others.

Ad Hoc. Most incident information sharing has traditionally occurred through Ad Hoc methods, such as email, instant messaging clients, and phone. Ad Hoc information-sharing mechanisms normally rely on an individual employee’s connections with employees in incident response teams of partner organizations. The employee uses these connections to manually share information with peers and coordinate with them to construct strategies for responding to an incident. Depending on the size of the organization, these Ad Hoc techniques may be the most cost-effective way of sharing information with partner organizations. However, due to the informal nature of Ad Hoc information sharing, it is not possible to guarantee that the information sharing processes will always operate. For example, if a particularly well-connected employee resigns from an incident response team, that team may temporarily lose the majority of information sharing channels it relies on to effectively coordinate with outside organizations. Ad Hoc information-sharing methods are also largely unstandardized in terms of what information is communicated and how that communication occurs. Because of the lack of standardization, they tend to require manual intervention and to be more resource-intensive to process than the alternative, partially automated methods. Whenever possible an organization should attempt to formalize its information-sharing strategies through formal agreements with partner organizations and technical mechanisms that will help to partially automate the sharing of information.

Partially Automated. Organizations should attempt to automate as much of the information sharing process as possible to make cross-organizational coordination efficient and cost-effective. In reality, it will not be possible to fully automate the sharing of all incident information, nor will it be desirable due to security and trust considerations. Organizations should attempt to achieve a balance of automated information sharing overlaid with human-centric processes for managing the information flow. When engineering automated information sharing solutions, organizations should first consider what types of information they will communicate with partners. The organization may want to construct a formal data dictionary enumerating all entities and the relationships between entities that they will wish to share. Once the organization understands the types of information they will share, it is necessary to construct formal, machine-processable models to capture this information. Wherever possible, an organization should use existing data exchange standards for representing the information they need to share.

The organization should work with its partner organizations when deciding on the data exchange models to ensure that the standards selected are compatible with the partner organization’s incident response systems. When selecting existing data exchange models, organizations may prefer to select multiple models that model different aspects of the incident response domain and then leverage these models in a modular fashion, communicating only the information needed at a specific decision point in the life cycle. In addition to selecting the data exchange models for sharing incident information, an organization must also work with its partner organizations to agree on the technical transport mechanisms for enabling the information exchange to occur in an automated fashion. These transport mechanisms include, at a minimum, the transport protocol for exchanging information, the architectural model for communicating with an information resource, and the applicable ports and domain names for accessing an information resource in a particular organization.

Security Considerations. There are several security considerations that incident response teams should consider when planning their information sharing. One is being able to designate who can see which pieces of incident information (e.g., protection of sensitive information). It may also be necessary to perform data sanitization or scrubbing to remove sensitive pieces of data from the incident information without disturbing the information on precursors, indicators, and other technical information. The incident response team should also ensure that the necessary measures are taken to protect information shared with the team by other organizations.

Granular Information Sharing. Organizations need to balance the benefits of information sharing with the drawbacks of sharing sensitive information, ideally sharing the necessary information and only the necessary information with the appropriate parties. Organizations can think of their incident information as being comprised of two types of information: business impact and technical. Business impact information is often shared in the context of a team-to-coordinating-team relationship, while technical information is often shared within all three types of coordination relationships.

Business Impact Information. Business impact information involves how the incident is affecting the organization in terms of mission impact, financial impact, etc. Such information, at least at a summary level, is often reported to higher-level coordinating incident response teams to communicate an estimate of the damage caused by the incident. Coordinating response teams may need this impact information to make decisions regarding the degree of assistance to provide to the reporting organization. A coordinating team may also use this information to make decisions relative to how a specific incident will affect other organizations in the community they represent. Coordinating teams may require member organizations to report on some degree of business impact information. In this case, for a hypothetical incident, an organization would report that it has a functional impact of medium, and information impact of none, and will require extended recoverability time. This high-level information would alert the coordinating team that the member organization requires some level of additional resources to recover from the incident. The coordinating team could then pursue additional communication with the member organization to determine how many resources are required as well as the type of resources based on the technical information provided about the incident. Business impact information is only useful for reporting to organizations that have some interest in ensuring the mission of the organization experiencing the incident. In many cases, incident response teams should avoid sharing business impact information with outside organizations unless there is a clear value proposition or formal reporting requirements. When sharing information with peer and partner organizations, incident response teams should focus on exchanging technical information.

Technical Information. There are many different types of technical indicators signifying the occurrence of an incident within an organization. These indicators originate from the variety of technical information associated with incidents, such as the hostnames and IP addresses of attacking hosts, samples of malware, precursors and indicators of similar incidents, and types of vulnerabilities exploited in an incident. While organizations gain value from collecting their own internal indicators, they may gain additional value from analyzing indicators received from partner organizations and sharing internal indicators for external analysis and use. If the organization receives external indicator data pertaining to an incident they have not seen, they can use that indicator data to identify the incident as it begins to occur. Similarly, an organization may use external indicator data to detect an ongoing incident that it was not aware of due to the lack of internal resources to capture the specific indicator data. Organizations may also benefit from sharing their internal indicator data with external organizations.

For example, if they share technical information pertaining to an incident they are experiencing, a partner organization may respond with a suggested remediation strategy for handling that incident. Organizations should share as much of this information as possible; however, there may be both security and liability reasons why an organization would not want to reveal the details of an exploited vulnerability. External indicators, such as the general characteristics of attacks and the identity of attacking hosts, are usually safe to share with others. Organizations should consider which types of technical information should or should not be shared with various parties, and then endeavor to share as much of the appropriate information as possible with other organizations. Technical indicator data is useful when it allows an organization to identify an actual incident. However, not all indicator data received from external sources will pertain to the organization receiving it. In some cases, this external data will generate false positives within the receiving organization’s network and may cause resources to be spent on nonexistent problems. Organizations participating in incident information sharing should have staff skilled in taking technical indicator information from sharing communities and disseminating that information throughout the enterprise, preferably in an automated way. Organizations should also attempt to ensure that they only share an indicator for which they have a relatively high level of confidence that it signifies an actual incident.

RECOMMENDATIONS

The key recommendations presented in this section for handling incidents are summarized below.

  • Plan incident coordination with external parties before incidents occur. Examples of external parties include other incident response teams, law enforcement agencies, Internet service providers, and constituents and customers. This planning helps ensure that all parties know their roles and that effective lines of communication are established.
  • Consult with the legal department before initiating any coordination efforts. There may be contracts or other agreements that need to be put into place before discussions occur.
  • Perform incident information sharing throughout the incident response life cycle. Information sharing is a key element of enabling coordination across organizations. Organizations should not wait until an incident has been fully resolved before sharing details of it with others.
  • Attempt to automate as much of the information sharing process as possible. This makes cross-organizational coordination efficient and cost-effective. Organizations should attempt to achieve a  balance of automated information sharing overlaid with human-centric processes for managing the information flow.
  • Balance the benefits of information sharing with the drawbacks of sharing sensitive information. Ideally, organizations should share the necessary information and only the necessary information with the appropriate parties. Business impact information is often shared in a team-to-coordinating team relationship, while technical information is often shared within all types of coordination relationships. When sharing information with peer and partner organizations, incident response teams should focus on exchanging technical information.
  • Share as much of the appropriate incident information as possible with other organizations.

Organizations should consider which types of technical information should or should not be shared with various parties. For example, external indicators, such as the general characteristics of attacks and the identity of attacking hosts, are usually safe to share with others, but there may be both security and liability reasons why an organization would not want to reveal the details of an exploited vulnerability.

Back to top button