**
In a groundbreaking move, a French start-up is leading the way in revolutionising how we understand public opinion. Naratis, founded by engineer Pierre Fontaine in 2025, is at the forefront of integrating artificial intelligence into the realm of political polling, promising faster, cheaper, and more insightful data collection than ever before. This innovative approach aims to reshape not only the methodology of polling but also how opinions are captured and analysed in a rapidly evolving political landscape.
The AI Advantage in Polling
Imagine a scenario where a young AI agent engages a respondent in conversation about their views on politicians. This is not a distant fantasy but a reality brought forth by Naratis. Unlike traditional polling methods that rely heavily on time-consuming interviews and small focus groups, Naratis employs AI to conduct qualitative research at an unprecedented scale. Fontaine highlights, “Our approach allows for deeper exploration of not just what people think, but how their opinions are formed and evolved.”
By automating the process, Naratis claims to complete studies in a fraction of the time—asserting that their method is “10 times faster, 10 times cheaper and 90% as accurate as human polling.” The implications of this are enormous: a poll that previously took weeks can now be executed in just a day or two, enabling clients to react to shifts in public sentiment almost in real time.
The Shift in Public Engagement
Polling has faced significant challenges over the years, particularly with declining response rates. According to AI consultant Stéphane Le Brun, the average response rate has plummeted from over 30% in the 1990s to below 5% today. This trend has led to increased costs and growing public scepticism regarding the accuracy of polls. Naratis’s model could be a game-changer, as it addresses these issues head-on by offering a more engaging way for respondents to share their views.
The conversational nature of AI polling allows respondents to express themselves freely, often leading to more candid answers—especially on sensitive topics. This is particularly relevant in France, where traditional polling methods have historically underestimated support for far-right parties.
Caution in the Age of Automation
Despite the clear benefits, the rise of AI in polling brings with it challenges that cannot be overlooked. Critics remain wary of the accuracy of AI-generated data. Fontaine acknowledges the shortcomings of quantitative polling but emphasises that qualitative research, which Naratis focuses on, is less about predicting outcomes and more about gaining insights into public sentiment.
Established firms in the polling industry are also experimenting with AI, incorporating it into market research and social media analysis. However, many remain cautious when it comes to political polling. Companies like Ipsos refuse to use AI-generated respondents in their surveys, prioritising human oversight to maintain trust and credibility.
The Future of Polling: A Hybrid Approach
The future of polling will likely be a hybrid model, combining the efficiency of AI with the essential human touch. While AI can significantly enhance data collection and analysis, human oversight remains crucial for interpreting results accurately and ethically. Fontaine’s vision for Naratis is not to replace human respondents but to transform how their voices are captured, turning traditional surveys into dynamic conversations.
As the polling industry navigates these changes, the challenge will be to find the right balance between automation and human involvement. The ultimate success of AI in polling will depend not just on the technology itself, but on how it is implemented, explained, and regulated to ensure transparency and trust.
Why it Matters
The integration of AI into opinion polling represents a significant shift in how we gauge public sentiment, with profound implications for political communication and engagement. As traditional methods falter under the weight of declining participation, the innovative practices of companies like Naratis could redefine the landscape, making it easier for voices to be heard and understood. However, the industry must tread carefully to ensure that while we embrace automation, we do not lose sight of the human experience behind the data. The future of opinion polling will be determined by our ability to navigate this delicate balance, shaping how societies understand and engage with the political arena.