Address Book Leeching

Address Book Leeching is a severe data-harvesting dark pattern formally categorized under “Privacy Manipulation” and “Forced Action.” While closely related to Friend Spam, Leeching focuses specifically on the covert extraction, upload, and permanent retention of a user's entire contact list. Applications often gate access to their service behind mandatory contact permissions, falsely implying the data will be processed locally or used solely for a transient feature. Instead, the application exfiltrates the address book to a remote server to build “shadow profiles” of non-users, map social graphs, or monetize the data through third-party brokerage. This pattern represents a fundamental violation of both user and non-user privacy, as it builds relational databases of individuals who have never interacted with or consented to the service.

Utility-Permission Decoupling

TorchMate — Flashlight

A local utility. Works entirely on your device.

Off

TorchMate requires contact access to continue

To use the flashlight, please allow TorchMate to read your contacts.

Condition 1: Utility-Permission Decoupling

State (U_{core}) = Blocked given P_{contacts} = False \land Dep (U_{core}, P_{contacts}) = \emptyset

Given

To identify the coercive nature of the data request, we define

U_{core}

as the primary, advertised utility of the application (e.g., a local music player or utility tool) and

P_{contacts}

as the operating system-level permission to read the address book. We apply a function

Dep (U, P)

to evaluate the technical dependency of the utility on the requested permission. The feature triggers if the application algorithmically blocks initialization or core functionality until the permission is granted, despite the mathematical absence of any functional requirement for contact data:

Visual Prominence of the Invite-All Affordance

SnapGrid — welcome

You’re in! Welcome to SnapGrid

Invite friends from your address book to make SnapGrid better for everyone.

Condition 2: Visual Prominence of the Invite-All Affordance

\frac{A ( N _{invite_all} )}{A ( N _{skip} )} > τ_{invite_dominance}

Given

To establish a visual baseline for Address Book Leeching, the algorithm identifies a mass-invitation button

N_{invite_all}

(e.g., “Invite All Contacts”) and a skip or deselect affordance

N_{skip}

. The feature triggers if the collective-invite button dominates the viewport with a bounding-box area exceeding the skip element by a dominance ratio

τ_{invite_dominance}

, pressuring the user toward indiscriminate contact sharing:

Semantic Framing of Contact-Sharing Consent

Find friends on SnapGrid

Find your friends on SnapGrid!

Let your contacts know you’re here! SnapGrid scans your address book to connect you with people you already know — no effort required.

Condition 3: Semantic Framing of Contact-Sharing Consent

Frame (T_{prompt}) = Benefactive \land Subject (T_{prompt}) = ThirdParty

Given

To establish a semantic baseline for Address Book Leeching, the algorithm inspects the text accompanying the contact-import permission prompt. The feature triggers if the prompt frames contact access as a social benefit to the user's friends (“Find your friends,” “Let your contacts know you're here”) rather than as a data-access permission, semantically reframing a privacy-invasive action as a social courtesy: