False Hierarchy
False Hierarchy is a structural and visual dark pattern formally categorized under “Interface Interference” and “Manipulating Visual Choice Architecture.” Distinct from general visual prominence, this pattern strictly requires a binary relational structure between two opposing actions. It intentionally assigns disproportionate visual prominence to the action that benefits the service provider (e.g., “Accept All Cookies”) while actively suppressing the visibility of the strictly opposing alternative action that benefits the user (e.g., “Reject All”). This creates a deceptive feedforward cue, overriding the user's conscious deliberation and inducing erroneous interaction with the highlighted element as if it were the mandatory path forward.
1
Structural Element Downgrading
Condition 1: Structural Element Downgrading
Given
Beyond aesthetic manipulation, interfaces frequently deprive user-favorable actions of their fundamental interaction signifiers and affordances. We define to represent the HTML tag type of an element, and as its computed clickable surface area padding. The feature triggers if the business action is rendered as a primary interactive component (e.g., a <button>), while its semantic opposite is structurally downgraded to a bare textual link (<a>). This is formalized by verifying the tag mismatch and confirming the padding mathematically approaches zero, effectively camouflaging the user's escape route:
2
Relational Visual Weight Disparity
Condition 2: Relational Visual Weight Disparity
Given
We define as the node representing the provider-favorable action, and as the opposing user-favorable action. To quantify their visual imbalance, we introduce a composite visual weight function , calculated as a linear combination of the vision-language model-extracted bounding box area , CSS background contrast , and font weight : The feature triggers if the visual weight of the business option vastly exceeds its direct semantic opposite. This indicates intentional relational suppression, formalized when the ratio pushes beyond a heuristic threshold :
3
Strict Semantic Opposition
Condition 3: Strict Semantic Opposition
Given
To establish this relational dependency, we evaluate two proximally close interactive DOM nodes, and , located within the same container . By extracting , representing the NLP-derived semantic intent of the text label for node , the algorithm evaluates their relationship. The feature cannot trigger unless the system first mathematically establishes that the two elements represent mutually exclusive, binary opposing choices: