French trust in search engines and AI grows, but data privacy concerns remain high, according to a 2026 Ipsos study.
Growing trust in AI responses
French people are increasingly trusting the responses given by search engines and generative AI. According to an Ipsos study conducted for Eskimoz in early 2026, 65% of 1,000 French participants found search engine content credible, up 9 points from the previous year, while 35% trusted generative AI, an increase of 8 points.Low trust in data protection
Despite this growing trust in AI responses, French people remain skeptical about how their personal data is managed. The study found that no platform received more than 35% trust regarding data protection, with generative AI trailing at 20%.Advert fatigue on video platforms
The survey also revealed that 59% of French people find advertising on video platforms bothersome, an increase of 7 points compared to the previous year. In contrast, discomfort with ads on search engines has decreased to 38%.Generative AI and advertising
Generative AI is less troubled by advertising, with only 29% of respondents finding it bothersome. This is largely because generative AI is still not heavily monetized through ads, unlike video platforms.The privacy paradox
Researchers Patricia Norberg, Daniel Horne, and David Horne identified the privacy paradox in 2007, noting a gap between people's stated privacy concerns and their actual behavior. As AI agents become more common, the nature of the exchange changes from consultation to delegation, requiring more personal information.E-commerce as a trust model
The history of e-commerce shows that trust can be built through infrastructure. In the early 2000s, online shopping faced resistance due to security concerns, but it gained traction with features like browser locks, secure payments, and return policies. Trust preceded widespread adoption.Priorities for brands
For brands, three main strategies are emerging: transparency, reassurance, and demonstration. This includes explaining data sources and recommendations, collecting less data, and providing guarantees and verified reviews to build trust in AI systems.
