Explanations of ISO/IEC 27005:2018
The source is ISO/IEC 27005:2018(en), third edition, July 2018. Title: Information technology — Security techniques — Information security…
Read articleISO/IEC 27005:2018
For: V, IA
Source: ISO/IEC 27005:2018(en), English. 8.3, Annex E informative
8.3.1: the analysis can be carried out at different levels of detail, depending on the criticality of the assets, on the extent of known vulnerabilities and on earlier incidents. It can be qualitative, quantitative or a combination. In practice, work is often qualitative first, in order to obtain a general indication of the risk level and of the major risks. Later a more precise or quantitative analysis of the major risks can be needed, because the qualitative analysis is usually less complex and less expensive. The form of the analysis should be consistent with the evaluation criteria from the context.
Qualitative: a scale of descriptive attributes for the magnitude of possible consequences, for example low, medium and high, and for the likelihood that these consequences occur. The advantage is understandability, the disadvantage the dependence on the subjective choice of the scale. Scales can be adapted. Different descriptions can apply to different risks. The qualitative analysis can serve as a first screen, where it is suitable for decisions, or where numbers or resources are not sufficient for a quantitative analysis. It should use factual information and data where they are available.
Quantitative: a scale with numerical values for consequences and likelihood, from various sources. The quality depends on the accuracy and completeness of the numbers and on the validity of the models. In most cases it uses historical incident data. The advantage is the direct relation to the information security objectives and concerns of the organization. The disadvantage is the lack of such data for new risks or weaknesses. Where verifiable data are lacking, a disadvantage can be the illusion of value and accuracy. How consequences and likelihood are expressed and combined into a risk level varies according to the kind of risk and the purpose of the result. Uncertainty and variation of consequences and likelihood should be taken into account in the analysis and communicated effectively.
8.3.2 Consequences. Input: the identified incident scenarios with threats, vulnerabilities, assets affected and consequences for assets and business processes. The business impact that can arise from possible or actual information security incidents should be assessed, taking into account the consequences of a breach such as loss of confidentiality, integrity or availability of the assets. After the identification of all assets under consideration, the values assigned to them should be taken into account in the assessment of the consequences.
One concept of business impact measures the consequences. The value can be expressed qualitatively or quantitatively. Any method that assigns a monetary value can generally provide more information for the decision. The valuation of the assets begins with their classification by criticality for the business objectives. The valuation is then determined with two measures: the replacement value, that is the cost of recovery, cleanup and replacement of the information, insofar as that is possible at all; and the business consequences of loss or compromise, for example adverse business, legal or regulatory consequences from disclosure, modification, unavailability or destruction of information and other information assets. This valuation can be determined from a business impact analysis. The value determined by the business consequence is usually distinctly higher than the simple replacement costs. The valuation is a key factor of the impact assessment of a scenario, because the incident can affect more than one asset or only a part. Different threats and vulnerabilities have different impacts. The assessment of the consequences therefore depends on the valuation of the assets on the basis of the business impact analysis.
Consequences can be determined by modelling the results of an event or a set of events, or by extrapolation from investigations or earlier data. They can be expressed in monetary, technical or human impact criteria or other criteria relevant to the organization. Sometimes more than one numerical value is needed, for different times, places, groups or situations. Consequences in time and money should be measured with the same approach as threat likelihood and vulnerability. Consistency is to be maintained in the quantitative or qualitative approach. “Is to be maintained”. Further information on valuation and impact assessment is in Annex B. Output: a list of assessed consequences of an incident scenario, expressed with reference to assets and impact criteria.
8.3.3 Likelihood. Input: the scenarios and the lists of existing and planned controls with effectiveness, implementation and use. The likelihood of the incident scenarios should be assessed. After the identification it is necessary to assess the likelihood of each scenario and of the impact occurring, qualitatively or quantitatively. “It is necessary”. That should take into account how often the threats occur and how easily the vulnerabilities can be exploited: experience and applicable statistics on the likelihood of the threat; for deliberate sources, motivation, capabilities that change over time, resources available, and the perception of attractiveness and vulnerability of the assets; for accidental sources, geographical factors, for example proximity to chemical or petroleum plants, the possibility of extreme weather conditions, and factors that can influence human error and equipment failure; vulnerabilities individually and together; existing controls and how effectively they reduce vulnerabilities. An example in the text: a system can be vulnerable to identity spoofing and to misuse of resources. The vulnerability to identity spoofing can be high because of a lack of authentication. The likelihood of misuse of resources can be low despite the lack of authentication, because the paths of misuse are limited. Depending on the needed accuracy, assets can be grouped or divided into elements. The same kind of asset can be exposed to other threats depending on the location, or existing controls can be differently effective. Output: the likelihood of the scenarios, quantitative or qualitative.
8.3.4 Level of risk. Input: scenarios with consequences and likelihood. The level of risk should be determined for all relevant incident scenarios. The analysis assigns values to likelihood and consequences, quantitatively or qualitatively. It is based on the assessed consequences and the likelihood. In addition it can take into account costs and benefits, the concerns of stakeholders and other variables, insofar as that is appropriate for the evaluation. The estimated risk is a combination of the likelihood of an incident scenario and its consequences. Examples of methods are in Annex E. Output: a list of risks with assigned value levels.
Annex E is informative. E.1 describes a high-level assessment. It allows priorities and a time ordering of the actions. For reasons such as the budget, not all controls can be implemented at the same time. It can be premature to begin a detailed risk management when implementation is planned only in one or two years. In order to reach the objective, the high level can begin with an assessment of the consequences, instead of with a systematic analysis of threats, vulnerabilities, assets and consequences. A further reason can be alignment with other plans, for example change or business continuity. Features that such an iteration can have: a more global view, in which technical aspects are considered separately from the business issues; a shorter list of threats and of vulnerabilities grouped by area, or a focus on risk or attack scenarios; risks often as general areas rather than as individual risks, and treatment that first proposes common controls; organizational and non-technical controls and management aspects of technical controls, or a few common technical protective measures such as backups and protection against malware. Advantages that the annex names: a simple first approach can encourage acceptance of the programme; it should be possible to gain a strategic picture; resources can be placed where they have the greatest benefit. The possible disadvantage: because the first analyses are at a high level and possibly less accurate, some processes or systems are perhaps not recognized for a second, detailed assessment. That can be avoided if sufficient information is available.
E.2 describes the detailed assessment: in-depth identification and valuation of assets, assessment of the threats to these assets and of the vulnerabilities. The results are used to assess the risks and then to determine the treatment. The detailed step generally needs considerable effort and is therefore most suitable for information systems with a high risk. “Can therefore be most suitable”. Consequences can be assessed quantitatively, qualitatively or in combination. For the likelihood, the period should be specified during which the asset has value or needs to be protected. The English wording is “needs to be protected”, inside the “should”. Many methods use tables and combine subjective and empirical measures. It is important that the organization use a method with which it is familiar, which it trusts and which produces repeatable results. Some examples of table-based techniques follow. For further techniques the annex refers to IEC 31010. That is not available. The examples use numbers to describe qualitative assessments. Users should know that further mathematical operations on numbers that are qualitative results can be invalid.
E.2.2, E.2.3 and E.2.4 are three examples: a matrix with predefined values, a ranking of threats by measures of risk, and the assessment of an asset for likelihood and possible consequences. The worked examples and the tables are not copied. A missing example is not a nonconformity. The introduction says the document provides no particular method.
Not in this text: a matrix, a monetary value as an obligation, and the three worked examples from Annex E. The evaluation against the criteria is 8.4, see 27005-06.