Full Disclosure, Less Trust? How the Level of Detail about AI Use in News Writing Affects Readers' Trust

Full Disclosure, Less Trust? How the Level of Detail about AI Use in News Writing Affects Readers' Trust

Source: cs.CY updates on arXiv.org — https://arxiv.org/abs/2601.09620 Date read: 2026-06-13 Connected to: none Escalation: store-only Escalation rationale:

What this is

A controlled factorial experiment (N=40) examining how varying levels of detail in AI disclosure statements affect reader trust in AI-assisted news articles. The work documents a "transparency dilemma" — the counterintuitive finding that detailed disclosure of AI involvement reduces trust compared to vague or absent disclosure.

What I took from it

This is a narrow applied study in the news domain, documenting a user perception effect rather than proposing a mechanism or generalizable law about protocolized systems. The core finding — that more information about AI use decreases trust — is empirically interesting but sits within existing HCI/sociology literature on transparency backlash and algorithm aversion. It does not theorize why detail triggers distrust (cognitive load? authority erosion? pattern-matching to deception cues?), nor does it test whether this effect generalizes to other domains (medical AI, financial systems, policy automation).

The work is useful as a data point confirming that naive "tell users everything" approaches fail, but it does not introduce a mechanism absent from the current inventory. It maps a phenomenon rather than explaining a law.

Research connections

• none currently (no established laws or active hypotheses to connect against yet)

Candidate laws or signals

CL-2601.09620-1: Disclosure granularity exhibits an inverted-U relationship with user trust in algorithmic systems — intermediate detail may optimize for perceived legitimacy, while full technical disclosure triggers skepticism or cognitive rejection.

(Low confidence; needs cross-domain replication and mechanistic investigation.)