According to Ernst & Young’s latest AI Sentiment Index 2025, only 24% of Americans take a close look at AI-generated content. This figure falls well short of the international average of 31%. Even though 74% of U.S. users appreciate the usefulness of these AI applications—slightly ahead of the global 73%—few actually question or tweak what the AI produces.
The survey surveyed more than 15,000 participants across 15 countries between December 2024 and February 2025. It also found that just 14% of U.S. citizens spend time fine-tuning or editing AI outputs, a rate that lags far behind the 32% seen in countries like China and India.
International comparisons paint an interesting picture. South Korea leads with a 42% review rate, while both China and India register 40%. Germany follows with a moderate 27%, slightly above the U.S., which shares similarly low engagement with nations like Sweden and France. This global snapshot was built using random stratified sampling with demographic quotas and enriched by 135 in-depth interviews exploring attitudes towards AI.
When it comes to healthcare, Americans remain noticeably cautious. Only 49% feel comfortable with AI predicting medical conditions from their data, trailing the global average of 57%. Likewise, a mere 31% are open to AI handling even basic medical consultations, while comfort levels in China and India hover around 61%.
Worries over AI-driven misinformation, including fake news and deepfakes, are common too, with 75% of Americans expressing concern—a sentiment mirrored globally. If you’ve ever felt uneasy about new technology, these figures might resonate with you, reminding us all to balance enthusiasm with a healthy dose of scepticism.
