When users access online services, they want to be confident that their data and services are secure and their privacy is protected. Institutions and Service Providers that offer online services also need to verify a user's identity to make sure only the right users are accessing the right information. That is why identity assurance is needed.
In general there are three types of authentication factors:
Strong authentication refers to the use of more than one of these factors. Generally this results in a higher level of assurance (LoA) about the user.
There are several international standards for identity assurance, like NIST (US), eIDAS (Europe, previously STORK) and ISO29115. SURFsecureID is based on ISO29115. The four levels of identity assurance commonly used are:
|LoA 1||Little or no confidence in the asserted identity|
|LoA 2||Some confidence in the asserted identity|
|LoA 3||High confidence in the asserted identity|
|LoA 4||Very high confidence in the asserted identity|
The different specifications elaborate on the meaning of these labels by specifying requirements for:
The resulting assurance level depends on the combination of these aspects. The aspect with the lowest score determines the overall assurance level (a chain is only as strong as its weakest link).
The required level of assurance can be estimated on:
These risks must be assessed to be able to decide what level of assurance is needed for your service (see also SURFnet guidelines).
The LoAs described by NIST and STORK primarily focus on the robustness of the authentication. The robustness of the technical infrastructure is mostly beyond their scope.
It is assumed that proper measures are taken to prevent authentication protocol threats such as eavesdropping, man-in-the-middle, replaying, and hijacking. Attacks are not limited to the authentication protocol itself. Other attacks include the use of malicious code to compromise authentication tokens, insider threats to compromise authentication tokens, social engineering to get a subscriber to reveal his password to the attacker, “shoulder-surfing”, fooling claimants into using an insecure protocol, when they think that they are using a secure protocol, or intentionally denying ever having registered by subscribers who deliberately compromise their tokens.
Other types of threats are (SAML) assertion related such as modification, disclosure, repudiation, reuse or redirect. Countermeasures should be taken to prevent these attacks as well. The most important ones are the use of digital signatures to sign assertions and the use of SSL/TLS to secure the communication channel.
Both control measures are required to fulfill the requirements for LoA2 and LoA3 and are already in place in SURFsecureID.
SURFsecureID solely focuses on authentication LoA. No LoA is assigned to the attributes of the user's identity.
Several attributes provided by the IdP (e.g. first and last name, e-mail address) will be validated during registration and identification. In theory a LoA could be assigned to these attributes, which in attribute-based access control scenario’s could make authorization more reliable. There are however some arguments against doing this:
Because of these arguments SURFsecureID solely focuses on authentication LoA.