LONDON, September 18, 2023 – After a year of hype and heat surrounding generative AI (GAI), a new Omdia survey finds that consumer uptake of GAI applications in key markets is still modest. However, engagement among existing users is high while more consumers have imminent plans to adopt GAI.
Omdia’s 2023 Consumer AI survey explores attitudes towards and usage of GAI among 3,000 plus people in the US, UK and China. The findings reveal that regular usage of GAI applications is still low, 10% overall across the three markets. These results may seem surprising given the intense supply-side activity, service/product launches and extensive media coverage. But noise and push does not automatically equate to consumer adoption and usage.
What is encouraging for generative service providers is that a further 17% of survey respondents have immediate plans to adopt GAI (i.e., within weeks), which should in theory see the survey sample user base rise to 27% in a short period.
Among those that have adopted GAI applications, current usage levels are healthy and are expected to increase. A quarter of respondents use GAI applications once a day (10%) or more (15%), while another quarter use GAI apps at least once a week, while 29% use such apps a couple of times per week. What is even more promising is that existing GAI users anticipate their consumption will increase over the next 12 months, and to a significant degree for 37% of respondents.
But a significant portion of consumers in the survey are resistant to GAI: 41% of respondents that have heard of GAI applications have no plans to use them. “These consumers may be deterred by the negative aspects of generative AI, which the survey shows are front of mind,” observes Eden Zoller, Omdia Chief Analyst, Consumer AI and study lead. The biggest issues for survey respondents are fears about how generative AI can disrupt job functions, be used to commit fraud, and/or manipulate people and spread misinformation (among other things). “Consumers needs to be shown that generative AI has value and can be trusted,” says Zoller.
