France's RN builds its own AI chatbot ahead of 2027
The Rassemblement National is developing a conversational AI tool trained on its own program, its leaders' speeches, and its parliamentary votes.
The goal: never again let a spokesperson go off-message.
At a Glance
France’s Rassemblement National (RN), the far-right party led by Jordan Bardella and longtime figurehead Marine Le Pen, is developing a proprietary AI chatbot to help its cadres and activists master the party’s 2027 presidential platform.
The initiative is a direct response to embarrassing messaging failures during the 2024 legislative elections, when a senior RN figure publicly defended a position the party had quietly dropped two years prior.
The tool builds on an existing system developed three years ago by Aurélien Lopez-Liguori, a former RN member of parliament for Hérault (2022–2026) and newly elected mayor of Agde, already used regularly inside the National Assembly.
This image is used for illustrative purposes only.
The next time a Rassemblement National spokesperson stumbles on a policy question during a live television appearance, artificial intelligence may be standing by. France’s leading far-right party is developing what France Inter, the French public broadcaster that first reported the story, described as its own “ChatGPT” — a conversational AI agent designed to give cadres and activists instant, on-message answers to questions about the party’s 2027 presidential platform.
“The objective is to allow our officials to better master our program.” [translated from French] — Jordan Bardella, president of the Rassemblement National
The tool and what feeds it
The system is still being built out. According to the party’s leadership, it is being trained on the presidential program — itself still a work in progress — as well as the speeches and op-eds of Marine Le Pen, RN’s longtime leader and three-time presidential candidate, and Bardella. The party’s voting record in the National Assembly, France’s lower house of parliament, is also being fed into the system.
The practical use case is straightforward: a party official preparing for a debate can ask the AI whether RN supports exempting businesses from employer social charges, or what the party’s current position is on family reunification, and receive an answer consistent with official doctrine.
Why 2024 still stings
The project has a specific origin story. During the 2024 legislative elections, Sébastien Chenu, a prominent RN member of parliament, went on C8, a French cable television channel, and defended the abolition of dual nationality — a position Marine Le Pen had actually abandoned two years earlier. The gaffe drew mockery from government ministers and exposed a structural weakness: in a party whose platform has shifted significantly as it pursues mainstream credibility, not all senior figures are keeping pace with the changes.
An RN official told France Inter the tool would allow the party to respond to attacks and mischaracterizations almost instantaneously. The underlying ambition is both defensive — preventing further off-script moments — and offensive, equipping the party’s media representatives with rapid-fire rebuttals.
AI as an instrument of partisan discipline
What RN is building goes beyond standard campaign preparation materials. It is an attempt at technological doctrinal enforcement — a challenge that is structurally acute for any mass political movement whose positions have evolved, sometimes sharply, in response to electoral calculations.
The 2024 episode illustrates a tension common to populist parties undergoing what French political analysts sometimes call dédiabolisation — a deliberate effort to shed the most extreme positions and broaden electoral appeal. That process doesn’t filter uniformly through a party’s ranks, and mid-level figures can remain attached to positions the leadership has quietly discarded. A centralized AI system trained on current official positions could theoretically close that gap — provided the underlying doctrine remains stable enough to be reliably encoded.
The approach also raises governance questions that are not unique to France. Who controls what enters the knowledge base? Who adjudicates when positions shift? By embedding doctrinal authority in an algorithmic system, RN would be creating a new instrument of internal centralization — one that may suit the party’s traditionally vertical structure, but could constrain the organic debates that sometimes produce programmatic renewal. Similar questions have emerged around AI-assisted campaign tools deployed by political parties in the United States and United Kingdom in recent electoral cycles, where the technology’s speed was seen as both an asset and a risk to genuine deliberation.
The tool has a direct predecessor: three years ago, Aurélien Lopez-Liguori, who served as an RN member of parliament for Hérault’s 7th constituency from 2022 to 2026 before being elected mayor of Agde, built a first-generation version of the system, which has since been used regularly by RN representatives inside the National Assembly chamber. The 2027 version appears to be a significantly more ambitious extension of that experiment.
The bottom line
Political parties have always issued briefing notes, talking points, and program summaries to discipline their public messaging. RN may be the first major French political party to attempt to automate that function through artificial intelligence. If the system performs as intended, it raises a broader question for every party on the political spectrum: in a world where a candidate’s official position on any issue can be retrieved in seconds, does political improvisation — with all its risks and occasional flashes of authenticity — have a future?
Sources: France Inter · France Info