CS4203 Flashcards
Asset Definition
Anything the computer system controls
How do buisnesses measure risk?
risk = expected cost x probability
Why does risk = expected cost x probability not work?
- cannot always estimate cost of something happening
- cannot always estimate probability
What is a qualitative way of measuring risk?
Create categories for hazards e.g. negligible, minor, major, survivable, existential.
What is a threat actor?
Could be company, state, hacker, etc.
What is a threat model?
Points of attack against a system. prioritised by risk, vulnerabilities and countermeasures.
What is an attack?
The execution of a threat using an attack vector by a threat actor.
What is a threat?
A way in which the system could be comprimised.
What is Identity?
Who you are.
What is an attack vector?
A specific vulnerability in the system that can be used in an attack against the system.
What is agency?
The things that you are allowed to do. (connected to identity)
What is privacy?
The requirement that information only be seen by those who have permission to see it.
What is authentication?
Proving a claim to an identity.
What is Anonymity?
Prevent something from being associated with an identity.
Can scripts act as users?
Yes and in fact it is very hard to distinguish between a person and a bot (sciprt acting as a user)
What is non-repudiation?
The inability to deny that you did something.
What is trust?
A combination fo agency and identity.
What is reputation?
Trust in an agent migrates from one system to another.
What is a vulnerability?
Aspects of a system that provide a mechanism for a threat.
What is a security model?
A precise statement of the rules that the system should implement and the properties it should maintain.
Why do you need a security model?
Security is a whole system issue, therefore everyone needs awareness of the risks, policies and warning signs.
List some consequences of being attacked?
- Locked out of systems
- Damage, time take to re-build and check integrity
- Reputation damage to the company
- Fines
- Lawsuits
- GDPR violation
Why can new technology be a bad idea?
- users will use it incorrectly unless they have training
- users will have to change their behaviour which they will resist
- extra security measures will not be liked and will be circumvented where possible
What are some criticisms of security models?
- might describe policies you cannot enforce
- cannot identify every threat this way
- cannot list every system component
- how do you introduce new components
- how do policies map to files and permissions?
- does the OD provide the needed service