[2Win5] Poster session 2

[2Win5-103] Public perceptions of AI’s risks and benefits

〇Satoko Yasuno¹, Mayu Terada², Hiroshi Nakagawa³, Yasutomo Kimura⁴, Tetsushi Tanibe⁵, Adam Randaii Smith⁶ (1.Chuo University, 2.Hitotsubashi University, 3.RIKEN, 4.Otaru University of Commerce, 5.Waseda University, 6.International Christian Univeristy)

Keywords:AI, public opinion on AI, knowledge on AI

_{How do we estimate the risk and benefit of AI? To what extent do we know about AI technologies? This report focuses on web based public opinion surveys conducted in 2023 and 2024, as a part of “The Dark Side of AI” project.
We conducted three surveys as follows: (1) The first wave of a general survey in December 2023(N=1377), (2)The second wave of general survey in December 2024(N=1200), (3) The professional survey in December 2024(N=500).
The main findings from the 2023 survey are: (1)Only 10-20% of respondents were aware of using various AI devices and services, (2)When respondents were shown seven terms such as “generative AI” and “deep learning”, 75% of respondents answered that they "don't know the meaning of any of them." (3)While the respondents’ knowledge on AI correlated with higher expectations for benefits of AI, there was no correlation between knowledge and concerns about AI's risks. (4)AI is perceived to judge fairer but less flexible. Other findings will be reported in the session.}