AI in the News: Does More Disclosure Build Trust or Hurt It?
What they studied
As newsrooms adopt AI, researchers tested how different disclosure levels affect readers' trust and behavior.
- 40 people read politics and lifestyle articles with no, one-line, or detailed AI disclosures; AI help was low or high.
- They measured trust, source-checking, and subscription choices, plus ran interviews.
What they found
- Trust dropped only with detailed disclosures.
- Source-checking rose for both one-line and detailed disclosures, strongest for detailed.
- Subscriptions tracked trust: they dipped when disclosures were very detailed.
- Topic interest drove fact-checking more than trust did.
- About two-thirds preferred detailed disclosures; others wanted a simple label with detail on demand.
Why it matters
Not every AI label causes a transparency dilemma. A light-touch notice plus optional deeper info may balance openness with trust.
Paper: https://arxiv.org/abs/2601.09620v1
Paper: https://arxiv.org/abs/2601.09620v1
Register: https://www.AiFeta.com
#AI #Journalism #Transparency #NewsTrust #HCI #UX #Media #AIethics
