According to a study conducted by Havas Market, more than 3 in 4 French people now use generative AI. Among them, 77% also use it to prepare their purchases. Enough to question the objectivity of the responses with the arrival of advertising.
More and more French people are using AI
While ChatGPT can now use your bank card to pay for your purchases, Havas Market has just revealed the results of its study entitled “French, AI & Commerce: uses & attitudes”. The latter shows that more and more French people are using AI on a daily basis. In October 2026, 77% of them say they use generative AI tools in the personal and/or professional sphere, compared to 66% in May 2026 and only 59% in April 2025.
This use of AI no longer only concerns the youngest. Thus, among those over 55, personal use of AI increased by 18 points compared to 10 points among 35-54 year olds and 5 points among 18-34 year olds. All age groups are now using these tools.
Increasingly frequent use of AI
The Havas Market study also looked at the frequency with which the French use AI. The results show that they use it more and more often. So, 7 in 10 users say they query AI tools at least once a week.
The uses are varied, but 77% of them have already used AI to prepare their purchases or obtain recommendations. More than half of them, or 54%, have already changed brands to follow a recommendation made by an AI rather than their first choice.
Users who trust AI
While artificial intelligence tools are the source of many concerns, the results of the Havas Market study show that the French have rather tendency to trust them. 50% of respondents would accept, for example, that AI chooses and buys a new product based on their tastes and general instructions.
On the other hand, while Google is banking on advertising in its AI to boost its revenue growth, the French say they are concerned about the potential influence of sponsored content and advertising in responses generated by AIs. 55% of users consider it essential or very important to know whether a recommendation made by an AI is sponsored or not. 36% also say that the presence of sponsored content would reduce their confidence in the neutrality of the responses received.

