The AI Dark Room: A World's Greatest Fear

A large Microsoft survey processed by Statista reveals widespread public concern regarding issues arising from artificial intelligence (AI).

71% of respondents expressed worry about AI-assisted fraud, although the specific types of fraud were not detailed. "It's highly likely this fraud is related to identity impersonation in public, with government officials, or close acquaintances of the respondents," wrote Florian Zand, Statista data journalist, on Wednesday, April 17, 2024.

Deepfakes—the falsification of photos, videos, audio, and other media to resemble a specific person—were a concern for 69% of respondents. This proportion is equal to the concern about online sexual harassment.

AI hallucinations were also a significant worry, selected by 66% of respondents. Statista defines this as chatbots, such as ChatGPT, providing nonsensical answers as data or facts in response to user queries.

Other AI-related concerns included data privacy breaches (62% of respondents) and the amplification of bias in information (60%).

"Overall, 87% of respondents expressed concern about at least one problematic AI scenario," wrote Florian.

The AI market is substantial, projected to be in the US$300-500 billion range in 2024. However, according to Statista, ignoring the potential dangers of AI could result in equally significant losses for society, impacting democracy, social structures, and individual lives.

This survey involved 16,795 respondents aged 13-64 across 17 countries, conducted between July and August 2023. The survey is included in the *Microsoft Global Online Safety Survey 2024* report.