Artificial intelligence is substantially transforming how pollsters collect public opinion, with a French emerging company called Naratis spearheading efforts into what promises to be a faster, cheaper alternative to conventional polling approaches. The company, established in 2025 by 28-year-old engineer Pierre Fontaine, deploys conversational AI agents to conduct detailed conversations with respondents, eliminating the labour-intensive process that has long characterised qualitative research. Rather than requiring respondents to select options, Naratis’s AI interacts with people in natural conversations designed to explore not just what they think, but how they think. The technology purports to provide results significantly quicker and at a tenth of the cost of conventional polling, whilst maintaining 90 per cent accuracy—a significant breakthrough as the polling industry grapples with declining participation levels and growing public distrust.
The Emergence of Dialogue-Based Polling
At the core of Naratis’s advancement lies a deceptively simple concept: substituting the transactional character of conventional polling with authentic dialogue. When a participant answers the phone, they meet a youthful, energetic AI voice asking open-ended questions about politics, society, and their personal views. Rather than simply recording answers, the system engages in real dialogue. Three distinct AI agents work simultaneously behind the scenes—one ensuring the respondent stays on topic, another seeking deeper insights when answers seem superficial, and a third confirming the person is genuine and not a bot exploiting the system. This layered approach transforms polling from a box-ticking exercise into something considerably nuanced and revealing.
The performance improvements are remarkable. Historically, qualitative research demanded lengthy and demanding work: assembling small groups of respondents, conducting individual interviews, documenting spoken exchanges, and then analysing responses for recurring themes and significance. Naratis collapses this timeline via what Fontaine terms “parallelisation”—numerous AI tools performing interviews at the same time rather than human interviewers working in sequence. A study that once took weeks and substantial sums of euros can now be completed within a day or two. Feedback frequently returns in a single day, enabling political movements, state institutions and entities to address unfolding events and evolving public sentiment with minimal delay, substantially altering the speed of polling work.
- AI agents carry out simultaneous interviews with multiple respondents
- Live analysis flags surface-level responses needing more thorough examination
- Fraud prevention prevents bot activity and dishonest responses from skewing data
- Results generated within hours instead of multiple weeks of traditional research
Speed and Efficiency Transform Research Surveys
The polling industry confronts an existential crisis. Response rates have plummeted from more than 30% in the 1990s to under 5% today, according to AI consultant Stéphane Le Brun. This sharp fall has created a downward spiral: fewer respondents mean increased expenses per completed survey, which in turn makes research less reflective of the broader population. Public trust in polling has eroded accordingly, with many viewing surveys as intrusive or unreliable. Against this backdrop, conversational polling powered by AI provides a lifeline, potentially reversing years of declining engagement by making the research process itself more appealing and participatory.
Naratis contends its AI-driven methodology achieves outcomes that are “10 times quicker, 10 times cheaper and 90% as precise as human polling.” These numbers, if validated independently, would represent a fundamental transformation in how organisations understand public opinion. The cost savings by themselves are transformative: a thorough qualitative investigation that previously demanded tens of thousands of euros and several weeks of work can now be completed for a fraction of the cost within days. This democratisation of access could enable smaller organisations, grassroots campaigns and community organisations to undertake thorough opinion research previously available only to well-resourced organisations.
Parallelisation: The Key Breakthrough
The innovation driving these gains is remarkably uncomplicated: parallel processing. Rather than human interviewers performing interviews in sequence—one conversation after another—AI agents operate in parallel across numerous respondents. This scaling of capacity without equivalent expense growth significantly changes the economics of polling. Where conventional research methods demanded considerable time and resources, AI-driven approaches compress timescales whilst lowering expenses, enabling companies to collect comprehensive, layered data on demand.
Accuracy Claims and Industry Scepticism
Naratis’s claim that its AI methodology attains 90% accuracy comparable to human polling has understandably prompted examination from experienced analysts. The polling industry, developed through decades of procedural improvement, remains wary of claims that automated systems can replicate the refined assessment of experienced human interviewers. Critics question whether conversational AI can genuinely identify the subtle social cues, hesitations and unspoken cues that skilled researchers use to explore more thoroughly respondent motivations. The company has failed to produce peer-reviewed studies validating its accuracy claims, leaving independent verification pending.
Beyond concerns about accuracy, sector analysts worry about possible prejudices embedded within AI systems themselves. If the algorithms underlying Naratis’s conversational agents are trained on skewed datasets or coded with unexamined assumptions, those flaws could consistently skew results across thousands of interviews. Additionally, respondents may change their conduct when interacting with machines rather than humans, either growing more forthright or more cautious based on their comfort with technology. These technical and psychological variables are largely unexamined territory, and their impact on polling reliability remains uncertain.
- Third-party assessment of accuracy claims is still pending from established research institutions
- Potential algorithmic biases could consistently skew results across extensive artificial intelligence survey programmes
- AI-human engagement dynamics may alter how respondents articulate authentic views and beliefs
The Synthetic Data Challenge
As AI polling grows, a troubling question surfaces: how will the public and regulators distinguish between genuine human responses and synthetic data generated by the very systems conducting the polls? The efficiency and speed that makes AI polling desirable also opens doors for manipulation. If an unscrupulous operator were to bolster actual responses with computer-generated data, the final dataset could seem statistically sound whilst having little in common to actual public opinion. The technology’s opacity compounds this risk—most voters would find it difficult to grasp how algorithms process and verify responses, making it difficult for them to trust the findings influencing political debate.
Naratis claims its systems feature fraud detection mechanisms, with one AI agent specifically assigned with identifying whether respondents are human or automated. However, this security feature itself is contingent on AI assessing AI, creating a self-referential flaw. As conversational systems develop greater complexity, differentiating real human exchanges from synthetic responses may prove technically unfeasible. The polling industry has traditionally maintained confidence among the public partly because its processes are fundamentally transparent—people provide responses, results are tallied. AI polling threatens to undermine that clarity, substituting human-readable processes with algorithmic black boxes that few can meaningfully audit.
Confidence and Compliance Concerns
Regulators across Europe are only now address AI’s involvement in opinion research and political polling. Currently, limited safeguards govern how AI systems collect, process and report polling data. Lacking strong oversight frameworks, the industry faces a credibility crisis if artificial information contaminates published results or if computational biases distort findings. French data protection regulators and the European Union’s AI Act implementation bodies must urgently create standards ensuring transparency, verifiability and responsibility in AI-enabled polling work before the technology becomes embedded in political processes.
The Combined Evolution of Consumer Insights
Despite the gains in efficiency AI polling offers, industry specialists suggest that human and machine-driven studies will probably coexist rather than one replacing the other entirely. Conventional polling approaches have endured decades of examination and remain integral to political institutions, regulatory frameworks and public understanding. Companies such as Naratis acknowledge that AI excels at speed and cost efficiency, yet human interviewers bring irreplaceable nuance—the capacity to detect fine emotional signals, adapt questions intuitively and build rapport that promotes candid responses. A balanced approach integrating both methods could produce deeper understanding whilst maintaining the openness voters increasingly expect from research shaping electoral discourse.
The move towards hybrid models, however, demands precise adjustment. Pollsters must set out definitive guidelines for when AI-gathered data should be weighted alongside conventional methods, and how findings should be presented to ensure the public understands which methods produced which conclusions. Preparing emerging researchers to work effectively with AI systems poses an additional obstacle, as does setting industry benchmarks that oversee the technology’s application. If handled with care, this evolution could revitalise opinion research by making it faster and more accessible whilst safeguarding the human judgment and ethical oversight that safeguard democratic discourse.