Consumers distrust AI but rarely check its answers
A new F-Secure survey indicates that while nearly 90% of AI users believe they are responsible for verifying AI-generated answers, 70% still do so only rarely.
A new survey by F-Secure among consumers in the US and UK reveals a dichotomy in how people interact with artificial intelligence. Nearly 90 percent of AI users state that it is their responsibility to verify the answers AI provides. However, a significant 70 percent report checking these answers only rarely, even when using AI to make high-stakes decisions.
The findings suggest that while consumers are aware of the potential for inaccuracies and biases in AI, this awareness does not consistently translate into user behavior. Possible reasons include an underlying trust in the technology, time constraints, or a default inclination to accept AI-generated information.
The survey also explored the types of high-stakes decisions being made with AI assistance. Although specific details were not provided, the mention of "high-stakes decisions" raises concerns about potentially incorrect advice in areas such as health, finance, or legal matters.
F-Secure's research highlights the need for greater public awareness regarding the risks associated with AI use and the development of tools or practices that encourage more critical evaluation by users. Responsible AI adoption requires active user judgment and verification, despite the technology's potential benefits.