An access analysis system obtains data about user requests to access particular applications, such as identifiers of the particular user and application involved, the time of the request, and (optionally) additional contextual data, and uses that data to generate user access distributions that quantify the distribution of a given user's requests to access applications over time. After one or more distributions have been generated for a particular user, when that user submits a new access request for an application, the access analysis system can compare the request to the previously-generated access distributions to determine whether (or to what degree) the request is anomalous. If the request is sufficiently non-anomalous, it can be granted with little or no additional actions required by the user or the user's device; if, however, the request is sufficiently anomalous, it can be denied, or additional information—such as additional user authentication factors—can be required.

Disclosed are example methods, systems, and devices that allow for generation and maintenance of a central identity databank for a user's digital life. The identity databank may include identity elements with payload values and metadata values corresponding immutable attributes of the user. A multifactor identity authentication protocol allows service provider devices to more reliably validate transactions with user devices via an identity system. The identity databank may include passwords, which may be generated by the identity system linked to user accounts and/or service providers. The passwords may be provided to service provider devices, eliminating the need for users to conceive of a multitude of varying passwords for the user's accounts.

Techniques are described herein for using special session identifiers to defer additional authentication steps (AAS) for at least some restricted application actions. A client session is associated with a special session identifier that is mapped to an authentication tier (AT) achieved for the session based on the satisfied authentication steps. Web servers that are enabled for AAS deferral include context information, which identifies a requested action, with session verification requests to an authentication service. The authentication service determines that AAS is required to perform an action when (a) the AT associated with the action is a higher-security tier than the AT associated with the session, or (b) the session is associated with an AT that is lower than the highest-security AT and there is no context information accompanying the request for session validation, in which case the authentication service assumes that the highest-security AT is required to perform the request.

The technology disclosed relates to authenticating users using a plurality of non-deterministic registration biometric inputs. During registration, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate sets of feature vectors. The non-deterministic biometric inputs can include a plurality of face images and a plurality of voice samples of a user. A characteristic identity vector for the user can be determined by averaging feature vectors. During authentication, a plurality of non-deterministic biometric inputs are given as input to a trained machine learning model to generate a set of authentication feature vectors. The sets of feature vectors are projected onto a surface of a hyper-sphere. The system can authenticate the user when a cosine distance between the authentication feature vector and a characteristic identity vector for the user is less than a pre-determined threshold.

A method including determining, by a first server, an encrypted authentication packet, the determining including, determining a crypted code field to indicate a type associated with the encryption authentication packet and that at least a portion of the encryption authentication packet is encrypted, and determining a crypted payload based at least in part on encrypting one or more fields of an initial authentication packet; and transmitting, by the first server to a second server, the encrypted authentication packet to enable the first server and the second server to conduct an authentication process. Various other aspects are contemplated.

Methods for increasing a data security of information shared and viewed within a social network platform is provided. Methods may include generating a social media impersonation score (“SMIS”) for a first user account on the social network platform. The SMIS may indicate a veracity of the user profile data associated with the first user account. Methods may include verifying the veracity of the user profile data by determining an authenticity of a corporate email address associated with the first user account. Methods may include determining that the corporate email address exists and verifying that a domain associated with the corporate email address is a trusted domain. Methods may include generating the SMIS based on results of the verifying and linking the SMIS to the first user account. Methods may further include displaying the first user account and the SMIS on a graphical user interface (“GUI”) of the social network platform.

A system and method for detection of email risk automatically is disclosed.

A server device receives, from a user device, a session initiation request and information identifying a location of the user device, and receives, from a monitoring device that is separate from the user device, an authentication request and information identifying a location of the monitoring device. The server device processes the session initiation request and the authentication request to authenticate a user of the user device, and determines, based on the location of the user device and the location of the monitoring device, that the user device and the monitoring device are collocated. The server device creates, after authenticating the user of the user device and determining that user device and the monitoring device are collocated, a session token, and sends the session token to the user device to enable the user device to access at least one resource of the server device.

Embodiments disclosed herein generally related to a system and method for assessing a fraud risk. In one embodiment, a method for assessing a fraud risk is disclosed herein. A web browser extension executing on the computing device identifies an account associated with the computing device. The web browser extension detects that the computing device navigated to a web page hosted by a third party server. The web browser extension determines that the third party server prompted the computing device to opt into two-factor authentication functionality. The web browser extension determines that the computing device did not opt into the two-factor authentication functionality. The web browser extension prompts, via an application programming interface (API), an organization computing system to update a fraud metric associated with the account.

An authentication model dynamically adjusts authentication factors required for access to a remote resource based on changes to a risk score for a user, a device, or some combination of these. For example, the authentication model may conditionally specify the number and type of authentication factors required by a user/device pair, and may dynamically alter authentication requirements based on changes to a current risk assessment for the user/device while the remote resource is in use.