Study reveals a double standard in how we view AI-assisted communication

A recent study published in the British Journal of Psychology uncovers a double standard when it comes to how people perceive the use of AI-assisted communication tools. People tend to believe others will employ these technologies more frequently than they themselves would, according to the research conducted by Université Paris Cité in Paris.

The study, which involved 1,242 participants from the UK and the United States, examined the acceptability and likelihood of both openly disclosed and secret AI use in various forms of communication, such as text, audio, and video. Participants rated their own and others' use on a five-point scale, with openly disclosed AI use scoring higher in acceptability (3.41 vs. 2.72 out of 5) and being expected to be used more often (3.73 vs. 2.80 out of 5).

However, when the two options were not presented side by side, the difference in acceptability became less pronounced, with open and secret use receiving average ratings of 56.87 and 53.62 out of 100, respectively. Despite this, the self-other difference remained significant, with participants estimating that others would be more likely to use the tools (71.41 out of 100) than themselves (40.44 out of 100).

Additionally, familiarity with AI communication tools was found to be positively correlated with higher acceptability and expectations of use. Conversely, concerns about AI-assisted communication, such as the potential loss of information or negative revelations about a user's character, led to less favorable ratings. The authors suggest that as these tools become more mainstream and people become more accustomed to them, concerns surrounding AI-assisted communication may lessen.