Changelog - August 14, 2026. Confidence intervals are now 90% Wilson score intervals computed on the effective sample size; the previous intervals (p ± t×SE, using the 90% critical value from a t distribution) could produce negative or zero-width bounds for small subgroups. The industry and occupation columns have been removed from the individual-level data download.

This polling was conducted by Epoch AI in partnership with Ipsos on the KnowledgePanel, a probability-based online panel recruited via address-based sampling. Probability-based sampling reduces self-selection bias and produces estimates that are more representative of US adults.

The survey was fielded March 3-5, 2026, and includes 2,021 respondents. A subset of the questions were asked only of participants who said they had used an AI service in the past week.

The full text of each option is available by clicking on the bars in the graphs above.

The results have been weighted to represent the adult population of the United States, ages 18 and older, using US Census benchmarks. We remove a small set of non-valid responses for the analysis.

The uncertainty intervals in the graphs are 90% Wilson score confidence intervals, computed on an effective sample size that accounts for unequal weighting. For each weighted proportion \(\hat{p}\) we compute

\(\frac{1}{1 + z^2 / n_\text{eff}} \left( \hat{p} + \frac{z^2}{2 n_\text{eff}} \pm z \sqrt{ \frac{\hat{p} (1 - \hat{p})}{n_\text{eff}} + \frac{z^2}{4 n_\text{eff}^2} } \right)\)

where \(z = 1.645\) is the 90% two-sided critical value from the standard normal distribution, and \(n_\text{eff}\) is the Kish effective sample size,

\(n_\text{eff} = \frac{\left( \sum_i w_i \right)^2}{\sum_i w_i^2}\)

computed over the respondents in the relevant base. Unlike the normal-approximation interval, the Wilson interval always lies within 0–100% and remains informative for proportions near 0% or 100%.

All estimates are weighted. The estimated proportion for any response category is:

\(\hat{p} = \frac{\sum_i w_i y_i}{\sum_i w_i}\)

where \(w_i\) is the weight for respondent \(i\) and \(y_i\) is a binary indicator (1 if the respondent selected the category, 0 otherwise).

Standard errors, reported in the downloadable data, are computed using Taylor series linearization to account for weighting. The Ipsos KnowledgePanel uses an unclustered and unstratified probability-based sampling design. The variance estimator therefore simplifies to:

\(\hat{V}(\hat{p}) = \frac{n}{n - 1} \cdot \frac{1}{(\sum_i w_i)^2} \cdot \sum_i \left[ w_i (y_i - \hat{p}) \right]^2\)

where \(n\) is the number of respondents in the relevant base.

To verify responses for providers who bundle AI features into other existing services (Google Gemini, Microsoft Copilot, Meta AI, Grok, and Perplexity), respondents who selected these products were shown a follow-up validation question asking how they accessed the service.

Thanks to Brad Edwards, Carlos Freitas, Scott Gardner, and Michael Sadowsky for helpful comments.