Artificial intelligence is substantially transforming how pollsters gather public opinion, with a French emerging company called Naratis spearheading efforts into what promises to be a faster, cheaper alternative to traditional survey methods. The company, established in 2025 by 28-year-old engineer Pierre Fontaine, utilises conversational AI agents to conduct detailed conversations with respondents, eliminating the labour-intensive process that has long characterised qualitative research. Rather than asking people to tick boxes, Naratis’s AI engages citizens in genuine dialogue intended to examine not just what they think, but how they think. The technology purports to provide results significantly quicker and at a fraction of the expense of traditional survey methods, whilst preserving 90 per cent accuracy—a significant breakthrough as the polling industry grapples with declining participation levels and growing public distrust.
The Rise of Conversational Polling
At the core of Naratis’s innovation lies a deceptively simple concept: substituting the transactional nature of traditional surveys with authentic dialogue. When a participant answers the phone, they meet a youthful, energetic AI voice asking open questions about politics, society, and their individual perspectives. Rather than simply recording answers, the system conducts genuine conversation. Three separate AI agents work simultaneously behind the scenes—one making sure the participant remains focused, another probing for further understanding when answers appear shallow, and a third confirming the person is authentic and not a bot gaming the system. This multi-layered strategy converts polling from a routine box-ticking task into something far more nuanced and revealing.
The productivity gains are impressive. Historically, qualitative research necessitated weeks of painstaking work: assembling small groups of respondents, performing individual interviews, documenting spoken exchanges, and then analysing responses for patterns and meaning. Naratis collapses this timeline using what Fontaine describes as “parallelisation”—multiple AI agents performing interviews at the same time rather than human interviewers working in sequence. A study that once consumed weeks and many thousands of euros can now be finished in one or two days. Data typically comes back in a single day, permitting campaigns, government bodies and groups to respond to breaking news and shifting public sentiment with minimal delay, radically transforming the pace of opinion research.
- AI agents conduct simultaneous interviews with numerous participants
- Instant analysis flags surface-level responses requiring more thorough examination
- Fraud prevention stops bot activity and dishonest responses from compromising data
- Results provided in just hours as opposed to weeks of traditional research
Velocity and Performance Reshape Survey Methodology
The survey sector faces an existential crisis. Response rates have collapsed from over 30% in the 1990s to under 5% today, according to AI consultant Stéphane Le Brun. This dramatic decline has generated a downward spiral: fewer respondents mean higher costs per completed survey, which in turn renders studies less representative of the broader population. Public trust in polling has eroded in turn, with many viewing surveys as intrusive or unreliable. Against this backdrop, AI-powered conversational polling offers a potential solution, potentially reversing decades of declining engagement by rendering the research process itself more engaging and participatory.
Naratis asserts its AI-driven approach achieves outcomes that are “10 times faster, 10 times cheaper and 90% as precise as traditional surveys.” These figures, if validated independently, would represent a fundamental transformation in the way organisations grasp public opinion. The financial savings by themselves are game-changing: a thorough qualitative investigation that once required tens of thousands of euros and several weeks of labour can now be completed for a fraction of the price within days. This broader accessibility could enable smaller entities, local campaigns and community organisations to undertake thorough opinion research formerly available only to well-funded institutions.
Parallelisation: The Key Breakthrough
The innovation driving these gains is elegantly straightforward: parallel processing. Rather than human interviewers carrying out interviews in sequence—one conversation after another—AI agents function concurrently across numerous respondents. This multiplication of effort without proportional increase in cost significantly changes the economics of polling. Where standard qualitative approaches required substantial commitment, AI-driven approaches shorten timelines whilst lowering expenses, permitting businesses to gather deep, nuanced insights on demand.
Precision Assertions and Sector Doubt
Naratis’s contention that its AI methodology attains 90% accuracy in line with human polling has understandably drawn criticism from recognised experts. The polling industry, developed through decades of technical advancement, remains cautious about claims that machine learning can replicate the subtle discernment of skilled researchers. Critics question whether conversational AI can genuinely identify the subtle social cues, hesitations and non-verbal signals that experienced practitioners use to probe deeper into respondent motivations. The company has failed to produce validated research confirming its accuracy claims, meaning independent verification remains incomplete.
Beyond accuracy concerns, sector analysts are concerned about possible prejudices embedded within AI systems themselves. If the algorithms powering Naratis’s conversational agents are developed using skewed datasets or coded with untested presumptions, those flaws could systematically distort results across thousands of interviews. Additionally, respondents may alter their behaviour when interacting with machines rather than humans, either becoming more candid or more guarded depending on their comfort with technology. These psychological and technical variables remain largely unexplored territory, and their effect on polling reliability remains uncertain.
- Third-party assessment of precision assertions is still pending from established research institutions
- Potential algorithmic biases could systematically distort results across large-scale AI polling operations
- Human-AI interaction effects may alter how respondents articulate authentic views and beliefs
The Synthetic Data Challenge
As AI polling scales up, a concerning question surfaces: how will regulators and the public distinguish between genuine human responses and synthetic data generated by the very systems performing the polls? The speed and efficiency that makes AI polling appealing also creates opportunities for distortion. If an unscrupulous operator were to supplement real responses with computer-generated data, the final dataset could look statistically solid whilst having little in common to actual voter sentiment. The algorithmic obscurity exacerbates the problem—most voters would have trouble comprehending how algorithms aggregate and authenticate responses, making it hard for them to have confidence in the results shaping political discourse.
Naratis asserts its systems incorporate fraud detection mechanisms, with one AI agent specifically assigned with determining if respondents are human or automated. However, this security feature itself depends on AI evaluating AI, generating a self-referential flaw. As interactive AI become increasingly sophisticated, separating authentic human dialogue from machine-produced answers may become technically impossible. The survey sector has long enjoyed public trust partly because its methods are conceptually straightforward—people provide responses, findings are compiled. AI polling risks compromising that transparency, replacing human-readable processes with inscrutable computational systems that few can meaningfully audit.
Confidence and Compliance Concerns
Regulators across Europe are just starting to grapple with AI’s place within opinion research and political polling. Currently, limited safeguards govern how AI systems process, analyse and disseminate polling data. Lacking strong regulatory frameworks, the industry faces a crisis of credibility if synthetic data infiltrates published results or if computational biases systematically skew findings. French data protection regulators and the EU’s AI Act enforcement bodies must without delay establish standards securing transparency, auditability and accountability in algorithmic polling processes before the technology becomes embedded in political processes.
The Combined Future of Opinion Research
Despite the gains in efficiency AI polling provides, industry specialists indicate that human and machine-driven studies will probably coexist rather than one displacing the other entirely. Traditional polling methods have weathered decades of examination and remain embedded in political institutions, regulatory frameworks and public understanding. Organisations like Naratis recognise that AI excels at speed and cost efficiency, yet human interviewers bring invaluable subtlety—the ability to read subtle emotional cues, adapt questions intuitively and build rapport that encourages candid responses. A measured strategy integrating both methods could yield richer insights whilst preserving the transparency voters increasingly demand from research influencing electoral discourse.
The shift to hybrid models, however, necessitates careful calibration. Pollsters must establish clear protocols for the circumstances under which AI data should be given weight alongside traditional responses, and the way conclusions should be shared to guarantee public comprehension of which methods produced which conclusions. Training a new generation of researchers to collaborate successfully with AI systems creates further difficulties, as does setting industry benchmarks that oversee the technology’s application. If approached strategically, this development could revitalise opinion research by making it faster and more accessible whilst maintaining the human judgment and ethical oversight that protect democratic discourse.