CSO — Managing consumer or citizen identities comes with two key problems--scale and cost--prompting organizations that require onboarding, authentication, and password management to look for ways to outsource this effort. Entertainment websites, online retailers, and even US federal government-to-citizen websites are experimenting with a federated model for more of their identity management life cycle. By using single sign-on (SSO) and attribute-sharing between "social" identity providers (IdPs) (i.e. Google (GOOG) and Facebook) and relying parties (RPs), this model effectively reduces cost and improves the customer experience.

Compared with a fully local model for consumer identity management, an RP can count on an IdP to provide initial and ongoing user identification, user password authentication with attendant password reset management, and automatic provisioning of the latest copy of key user attributes at each login.

What consumer-scale IdPs dont do, however, is assure that each identity maps to a real-world person.

[Also see Facebook adds two-factor authentication]

Assurance involves two factors: 1) During credential issuance, were the identity verification measures that bound the issued credential to a real-world person accurate? and 2) when the credential is presented at runtime, how strong are the authentication and other protection measures that continue to bind the "right person" to it? While closed networks, such as SAFE-BioPharma and Surescripts in the medical world, provide identity assurance to their communities through relatively heavyweight PKI mechanisms, the popular free consumer IdPs today offer neither meaningful assurance promises nor the technical underpinnings to support reliance on them.

Not surprisingly, many consumer RP scenarios would benefit from ensuring at login that the user identities correspond to real people for a variety of liability, forensic and business agility reasons. Having this assurance could help with identity issues ranging from preventing criminal gangs and terrorist groups from activating SIM cards, to maximizing the security of online auctions, to offering better "verified accounts" for social networking sites.

Forrester's take: Over the next three years, advances in consumer federated identity and identity assurance will provide new options for distributing risk, liability, and cost in managing identities. In the meantime, the good news is that the U.S. federal government is at the forefront of a variety of innovations in federated identity assurance, for both government-specific and commercial scenarios that require trusted online interactions. In addition to web 2.0 solutions, security professionals should leverage the following standards to address the challenges of outsourcing identity assurance.

NIST Assurance Levels Offer Common Ground

In 2003, the US Office of Management and Budget (OMB) issued Memo 04-04, which maps a system of risk assessments to four broad assurance levels. Building on this memo, the US National Institute of Standards and Technologys (NIST) Special Publication 800-63 defines a standard for credential issuance, identity proofing, and authentication, giving operational meaning to each level. Level 1, for example, is used to guarantee baseline "security hygiene" in lightweight, privacy sensitive scenarios, while Level 4 is reserved for non-consumer-scale applications requiring high levels of security.

Continue Reading