True costs of misinformation| mountains of evidence: Processual “redpilling” as a Socio-technical effect of disinformation

Summary

Despite the title referencing “redpilling” and disinformation, the structured summary describes a paper that bridges HCI dark patterns scholarship with EU platform regulation. The authors systematically map 59 dark patterns from Gray et al.’s unified ontology onto the three user-autonomy violations prohibited under Article 25 of the EU Digital Services Act—deception, manipulation, and distortion/impairment. Through iterative qualitative coding, they derive eight design factors (organized into an “Information Space” and a “Choice Space”) that explain why a given pattern constitutes a particular violation. The result is a “law-to-design” framework intended to support both regulator enforcement and compliance-by-design, demonstrated on attention-capture damaging patterns and an ongoing EU Commission case against X’s paid blue checkmarks.

Key Contributions

• A complete mapping of 59 meso- and low-level dark patterns to DSA Article 25’s three autonomy violation categories, with explicit per-pattern reasoning.
• An eight-factor design vocabulary (Information Availability, Correctness, Framing, Presentation; Choice Simplification, Presentation, Availability, Effort) operationalizing when a pattern violates autonomy.
• A reverse “law-to-design” framework complementing prior “design-to-law” work.
• Demonstrations of extensibility: application to 11 Attention Capture Damaging Patterns and an in-situ analysis of X’s blue checkmarks under EU enforcement.
• A shared interdisciplinary vocabulary linking HCI design analysis with legal/regulatory practice.

Methods

Two authors independently coded 59 dark patterns in AirTable using Santos et al.’s legal interpretation of Article 25, resolving disagreements over three structured rounds. Design factors were derived inductively from rationale memos and then applied deductively in a second coding pass on 34 low-level patterns (Cohen’s κ = 0.679). Co-authors validated the final rationales. The framework was then stress-tested on attention-capture patterns outside the original ontology and on a real EU Commission enforcement case.

Findings

• Only 17 of 59 patterns map to a single autonomy violation; 31 implicate two violations and 11 implicate all three.
• High-level categories show strong internal consistency: all Sneaking patterns are deceptive, all Obstruction patterns are distorting/impairing, and all Social Engineering patterns are manipulative.
• Forced Action is the most heterogeneous category; Interface Interference is dominantly manipulative with deception as secondary.
• Deception + manipulation is the most common pairing (22 patterns); manipulation + distortion is rarest (2 patterns).
• Deception and manipulation behave as a continuum of influence severity, while distortion/impairment is conceptually distinct.
• Violations can co-occur in alternative, sequential (e.g., drip pricing), or reinforcing (e.g., Sneak into Basket) configurations.

Connections

No other papers have been provided under shared topics, so there are no register neighbours to link. Intellectually, the work sits squarely between Gray et al.’s unified dark patterns ontology and Santos et al.’s legal reading of DSA Article 25, and it contributes to the emerging area of regulatory design auditing that translates between HCI and platform law.

Podcast

A research-radio episode discusses this paper: Listen