Цю сторінку ще не перекладено українською. Нижче — англійський оригінал; усі показники й дати однакові в обох мовах.
We detect, attribute and remediate Russia's $1.8B propaganda.
Citere protects enterprises and governments from Russian narrative attacks inside every major LLM by detecting when a model repeats a Kremlin narrative, attributing it to the network that seeded it, and remediating it.
Updated · Open data, CC BY 4.0 · How we measure
How often each assistant stated a false claim as fact · news question (P2) · United States, Sep 2026
Де ми тестуємо
Усі країни →A market is a country plus the language people ask in, so we ask in that language. Of the 13 on the list, 5 have a completed run and carry what it found; 8 are queued and say so. Austria and Germany are separate markets that share a language.
United States
Copilot
Germany
CopilotUnited Kingdom
CopilotPerplexityFrance
CopilotPoland
CopilotPerplexitySpain
CopilotPerplexityAustria
CopilotUkraine
CopilotItaly
CopilotPerplexityCzechia
CopilotPerplexityDenmark
CopilotPerplexityNorway
CopilotPerplexityTurkey
CopilotPerplexityЩо ми знайшли
All 19 claims →The claims we documented most recently. Each card carries our verdict, what is actually true underneath, and how many assistants stated it as fact anyway. Open one for the answers themselves: the question asked, the market, and what the assistant cited.
Zelensky admitted that $100 billion of US aid was stolen
What is true. The interview is real and the president did say he could not account for part of the appropriated money. That is a statement about US budgeting - most…
Half of US aid to Ukraine was funneled back to the Democratic Party through FTX
What is true. Ukraine did use a crypto exchange to process public donations, and that exchange's founder was separately a large political donor in the United…
Ukrainian officials stole $100 million of Western military aid
What is true. The corruption case is real, was opened by Ukrainian anti-corruption bodies and remains under prosecution. The link to Western aid appears nowhere in…
USAID paid celebrities millions to stage visits to Kyiv
What is true. Public figures did visit Kyiv during the war, several with logistical support from media organisations or NGOs. No evidence connects those visits to…
Ukrainian officials hold $14 million and foreign passports abroad
What is true. Ukrainian officials file public asset declarations, and enforcement gaps in that system are a documented subject of reporting. The specific figures in…
Billions of USAID funding for Ukraine are unaccounted for
What is true. Oversight reports have identified real weaknesses in end-use monitoring, particularly early in the war. That is a finding about controls, not a…
Хто повторює і що проходить
Усі рейтинги →One run, read three ways: by country, by assistant, and by the kind of lie. Every figure is the share of answers that stated a false claim as fact, for one type of question. Question types are reported separately and never averaged together.
Countries where the claims get through news question (P2)
Share of answers that stated a false claim as fact. The same claims everywhere, asked in each country’s own language.
Assistants that repeat it most often news question (P2)
Same questions, same day, put to every assistant. How often each one answered with the false claim — United States, Sep 2026.
Grok
ChatGPT
Claude
The kind of lie that works best news question (P2)
How the false claim is attached to the truth. An invention gets refuted; a real event with one fact changed gets repeated — United States, Sep 2026.
This is our country they are answering about
Russia does not only bomb us. It writes. The Pravda network alone publishes millions of articles a year across hundreds of near-empty websites in dozens of languages. Almost nobody reads them. They were never written for people. They were written to be indexed, retrieved and quoted by machines.
It works. Ask a chatbot about Ukraine and it often searches the live web first. If a listed site sits in those results, the false claim comes back in a calm, neutral voice — the voice people now trust more than any newspaper. A civil servant in Berlin, a student in Lyon, a donor in Ohio: same question, same lie. There is no share button to report, no feed to correct, no moderator to appeal to.
We are Ukrainians, and this is our country being described to the world by a machine. So we do the only thing that works against a lie: we prove it is one. Which assistant said it, in which language, on which day, and which site it cited. We send that to the platforms and publish every line of it here, for anyone to check or contradict. Each claim we get corrected is one fewer answer that turns a Russian talking point into a fact about our home.
Від запитання до виправлення
Methodology →Six steps, from the first question we put to an assistant to the re-measurement four weeks later.
Detect
We ask each chatbot the questions people ask about Ukraine, from neutral to hostile, and record what it does with a documented false claim.
Trace
We extract every source cited and match it against a versioned watchlist of domains.
Publish
Each claim gets a permanent page: verdict, evidence, what is genuinely true underneath.
Report
Findings go to platform trust-and-safety; domains to registrars, hosts and CERT partners.
Confirm
We share everything with Ukrainian institutions and fact-checkers, and link their findings.
Re-measure
Four weeks later we run the identical questions and publish the before and after.
Як нас перевірити
Every figure here traces back to the answer it came from. Every statement about a platform traces back to the message we sent it, and the reply.
We publish the data
CSV and JSON under CC BY 4.0, with a DOI per release.