tech
‘Our assumptions are broken’: how fraudulent church data revealed AI’s threat to polling
Experts say paid participants are using automated tools to generate unreliable survey responses at scale
TL;DR
- A Bible Society report claiming a resurgence in church attendance in Britain, particularly among young people, was based on fraudulent data and has been withdrawn.
- Researchers warn that online opt-in surveys are increasingly compromised by bogus data, often generated by paid participants using AI tools to complete questionnaires rapidly.
- This issue raises concerns about the reliability of survey research and its potential to shape public discourse based on manipulated findings.
- Experts like David Voas and Sean Westwood highlight that AI models can be weaponized to systematically bias survey results, making manipulation difficult to detect.
- The problem is exacerbated by the speed of AI development, making detection methods quickly obsolete, and the tendency for bogus respondents, particularly younger ones, to exhibit 'positivity bias'.