Generative AI has become a fully embedded habit for billions worldwide. In the US, 73% of the consumers we surveyed in April said they were using generative AI applications and tools in their personal lives, up from 53% in September 2025. Adoption in Germany, France, and the UK has advanced with similar rapidity, with the average personal usage rate rising from 47% to 69% over the same period.

Crucially, generative AI use is a habit that builds over time. In the US, about 50% of early adopters—those who started using generative AI more than two years previously—told us they were using it once a day or more. Of those who started using it in the preceding year, slightly less than 30% were daily users.

However, there’s also a paradox at the heart of generative AI’s dash into the consumer mainstream: Trust in the technology’s accuracy has been falling in the US, even among the early adopters who are likely to be its most sophisticated and pragmatic users. Our research illustrates this contrary trend.

The worries about accuracy vary greatly with the nature of the task that AI is carrying out. For instance, consumers in both the US and Europe said in April that they had slightly more trust in the accuracy of generative AI when using it for learning than they did in September 2025. Consumers tended to say they were more mistrustful of AI’s accuracy when it was deployed in high-stakes activities, such as the provision of financial advice. That suggests consumers are displaying some calibrated skepticism.

When it comes to generative AI’s likely impact on employment, Gen Z once again stands out. These young adults, almost all of whom are under the age of 30, worry most about what AI will do to their jobs and earning power, particularly in the US. That’s unsurprising, given that many members of Gen Z are still finding their footing on the career ladder.

Consumer confidence in AI may well be further dented by fears about the longer-term safety and social impact of frontier AI that were publicly voiced by the creators of leading AI platforms in September 2026. Over the coming months, however, companies seem unlikely to pause the rollout of non-frontier AI technology, particularly for applications that deliver proven productivity benefits.

As deployment continues, our findings suggest that organizations could benefit from giving employees as much agency as possible in their adoption of generative AI, rather than forcing it upon them. Consumers report being less worried about the employment impact of AI if their use of it is intrinsically motivated (to experiment or delegate boring tasks, say) rather than arising from external pressure, such as an obligation to use AI at work.

For executive teams, there’s work ahead to manage consumer and employee concerns about AI, particularly with Gen Z. But a questioning, engaged public may be just what AI needs to realize its full potential after entering the mainstream so emphatically.