A Model for New Zealands Identity Verification Service

  • Slides: 14
Download presentation
A Model for New Zealand’s Identity Verification Service Prof. Clark Thomborson Trust 2008 Villach,

A Model for New Zealand’s Identity Verification Service Prof. Clark Thomborson Trust 2008 Villach, Austria 11 th March 2008

Objectives of this model n To study identification in a hierarchy, in a peerage,

Objectives of this model n To study identification in a hierarchy, in a peerage, and in an aliased relationship. n n To develop a simple yet meaningful model of a complex identification management system. n n I believe these three types of relationships are of crucial importance in trusted computing. Case study: New Zealand’s Identification Verification System To distinguish three types of authentication: n “What you have”, “What you are”, and “What you know”. NZ IVS March 2008 2

The CEO Model 1. Credentials 2. Entities 3. Observations n n n An entity

The CEO Model 1. Credentials 2. Entities 3. Observations n n n An entity can disclose a self-signed credential, revealing its signing key. Entities can observe their inferiors and aliases. Entities reveal information about other entities to their superiors and aliases. NZ IVS March 2008 3

Three Types of Relationships Hierarchical: A superior observes its inferiors. An inferior entity discloses

Three Types of Relationships Hierarchical: A superior observes its inferiors. An inferior entity discloses “everything” to their superior. A superior discloses only its signing key to its inferiors. n n This is a Bell – La Padula system. Superiors do not disclose to inferiors. Inferiors can not observe superiors. Peering: A peer does not disclose their personal identifiers to other peers. Peers cannot observe each other. n n n Peers share an pseudonym: their Peer Identifier. Peers can vote anonymously. Organisations (whether hierarchical or peering) can communicate via any entity which has an alias in each organisation. n n n Aliased entities managed the “bridging relationships” defined in my previous writing. Aliases are observable. Our aliases will publish credentials to their superiors, and are thus a security risk to a confidential hierarchy. NZ IVS March 2008 4

The Hierarchy n n Information flows King, President, Chief upwards, toward the Justice, Pope,

The Hierarchy n n Information flows King, President, Chief upwards, toward the Justice, Pope, or … most powerful actor (at the root). Commands and trust flow downwards. The King is the most Peons, illegal immigrants, felons, privileged. excommunicants, or … The peons are n Information flowing up is “privileged”: the most trusted. allowed by default, might be required. n Information flowing down is “trusted”: prohibited by default. n See TCSEC (Orange Book): LOCKix. NZ IVS March 2008 5

The Alias (in an email use case) n n n We use “aliases” every

The Alias (in an email use case) n n n We use “aliases” every time we send personal email from our work computer. We have a different alias in each organisation. We must be careful what information we disclose about our organisations, and what information we introduce into our organisations. Agency X Hotmail C, acting as a governmental agent n C, acting as a hotmail client An aliased relationship is bidirectional trusted. Each of our aliases is in a different security environment. NZ IVS March 2008 6

Another Example: B 2 B e-commerce n n n Use case: employee C of

Another Example: B 2 B e-commerce n n n Use case: employee C of X purchasing supplies through employee V of Y. Company X Employee C creates a hotmail account for a “purchasing” persona. Purchaser C should reveal minimal information to V. C, acting as an employee • • Company Y Employee V C, acting as a purchaser Most workflow systems have rigid personae definitions (= role assignments). Current operating systems offer very little support for aliases, other than virtualisation. Important future work! NZ IVS March 2008 7

The Peerage n n n Information flows upwards by default (“privileged”). Commands and trust

The Peerage n n n Information flows upwards by default (“privileged”). Commands and trust flow downwards. Downward information flows are “trusted” (filtered, audited, etc. ) A peerage can be modeled by Bell-La Padula, because there is a partial order on the actors’ privileges. Peers can vote anonymously. Peers, Group members, Citizens of an ideal democracy, … Facilitator, Moderator, Democratic Leader, … n Equality of privilege is the default in a peerage, whereas inequality of privilege is the default in a hierarchy. NZ IVS March 2008 8

Example: A Peerage Exerting Audit Control on a Hierarchy OS Root Administrator Auditor Users/

Example: A Peerage Exerting Audit Control on a Hierarchy OS Root Administrator Auditor Users/ Peers IG 1 IG 2 Inspector-General (an elected officer) Chair of User Assurance Group NZ IVS March 2008 • Peers vote on their officers, including at least one Inspector-General. • The OS Administrator makes a Trusting appointment when granting auditor-level Privilege to an alias of an Inspector-General. • The Auditor should disclose a report to their Inspector-General alias. • The audit report can be read by any Peer. • Peers might disclose the report to non-Peers. 9

Application to Identity Management n The system on the preceding slide has very complex

Application to Identity Management n The system on the preceding slide has very complex security objectives. n n Each entity has complex capabilities, privileges, and trusting relationships with other entities. See the ECO extension of Secure Tropos (Giorgini et al. , 2006). Roughly: “privilege” = “entitlement”. I don’t want to get bogged down on TC UAGs here, but let’s talk offline if you want to help me create one! The CEO model is greatly simplified. n CEO entities have very simple privileges, trusts, capabilities, and objectives – suitable only for identity management systems. NZ IVS March 2008 10

Observations in the CEO model Hierarchical: a superior can observe its inferiors, but an

Observations in the CEO model Hierarchical: a superior can observe its inferiors, but an inferior cannot observe its superior. n n n A superior has just one privilege: it can see the “observable identity” (Ob. Id) of all inferiors. Application: a surveillance camera in a workplace can collect biometric data on all employees and visitors. Peering: a peer cannot observe its peers. n n Peers share a Peer Identity (Peer. Id) but are otherwise anonymous. There are two types of entities. n n n Active entities can observe the Ob. Id of all their aliases, and will disclose this information (in a credential) to their superior. Passive entities can be observed, but cannot disclose credentials. Application: our identification tokens are active aliases. Application: any database record that contains one of our identifiers is a passive alias. NZ IVS March 2008 11

Disclosures in the CEO model Hierarchical: inferiors disclose their Peer. Id to their superior.

Disclosures in the CEO model Hierarchical: inferiors disclose their Peer. Id to their superior. Alias: active entities observe the Ob. Id of all their aliases. Alias: active entities disclose “everything they know” to their other aliases. Hierarchical: inferiors disclose “everything they know” to their superior. n n A CEO model of an identification management system n n will help us explain it; will help us figure out “who learns what” (identification vs. preservation of anonymity); will help us prove that its credentials guarantee a uniqueness property. NZ IVS March 2008 12

NZ’s IVS Crown DIA IVS Referee Citizen DIA Id VID for SA 1 IVS

NZ’s IVS Crown DIA IVS Referee Citizen DIA Id VID for SA 1 IVS VID for SA 2 SA 1 SA 2 GLS VID at SA 1 • Anonymised IDs can be created. • Session IDs are transient. igovt VID SID for SA 1 • A citizen can have at most one Verified ID (VID) at each agency. igovt UVID SID for SA 2 VID at SA 2 SID for SA 2 Anon at SA 2

Did I meet my objectives? n To study identification in a hierarchy, in a

Did I meet my objectives? n To study identification in a hierarchy, in a peerage, and in an aliased relationship. n n To develop a simple yet meaningful model of a complex identification management system. n n Do you understand these relationships? (Do you agree that they are crucial in trusted computing? ) Do you understand how New Zealand’s Identification Verification System would provide semi-anonymous and semi-unique credentials? To distinguish three types of authentication. n Do you believe that the CEO model could distinguish “What you have”, “What you are”, and “What you know”? NZ IVS March 2008 14