CivilNet Check (#CivilNetCheck) is CivilNet’s fact-checking unit, operating since December 2021.
Founded in 2011, CivilNet is one of the leading independent media outlets in Armenia. Since its inception, CivilNet has provided multi-faceted, evidence-based news and analysis. The establishment of CivilNet’s fact-checking unit as a separate department was motivated by the imperative to counter the flow of misinformation, fake news and manipulation, and to present the truth to the public.
#CivilNetCheck’s main mission is to combat fake news in Armenia, uncover and inform the public about misinformation and manipulation, and promote fact-checking journalism.
In March 2023, #CivilNetCheck became a member of the International Fact-Checking Network (IFCN). It conforms its activity and methodology to the standards of the American Poynter Institute of Journalism. We are guided by the fact-checking principles of the International Fact-Checking Network (IFCN):
Commitment to impartiality and fairness
Commitment to transparency in sourcing
Commitment to transparency about the Foundation and its funding
Commitment to transparency of methodology
Commitment to making corrections in an open and honest manner
In July 2024, #CivilNetCheck started working with Meta, which operates Facebook and Instagram, becoming the first Armenian news organization to join Meta’s Third-Party Fact-Checking Program. In that capacity, #CivilNetCheck employs native Armenian language fact-checkers to review and rate the accuracy of social media content.
In February 2025, #CivilNetCheck joined the European Fact-Checking Standards Network (EFCSN) as a member, becoming the first organization in Armenia to do so. This membership attests to #CivilNetCheck’s compliance with the professional and ethical standards established by EFCSN, which include independence, transparency, methodological clarity, and accountability.
#CivilNetCheck structure
#CivilNetCheck consists of a team of responsible and professional journalists, which is continuously expanding.
The editor of #CivilNetCheck is Ani Grigoryan, whose responsibilities include selecting and approving topics and materials, checking and editing journalists’ articles, and rechecking the facts included in them, as well as writing articles. Leading #CivilNetCheck, she is also responsible for keeping the department up-to-date, implementing new tools, and meeting international standards of fact-checking journalism. She reports to the CivilNet editor-in-chief.
The fact-checking journalists of #CivilNetCheck: Shushan Stepanyan, Armen Mkrtchyan, Nelli Lazaryan, and Hayk Hovhannisyan, are the main authors of fact-checking materials. They monitor the media, social networks, and the general internet, follow the speeches and interviews of political and public figures, identify statements and claims subject to verification, carry out their verification, and, after coordinating with the editor, prepare articles and materials about them. The journalists are accountable to the editor and the editor-in-chief.
In addition, #CivilNetCheck collaborates with independent fact-checking journalists. The editor of #CivilnetCheck manages the work with them.
Videos and visual content published by #CivilNetCheck are produced with the help of CivilNet’s video production team. Articles and materials are published and distributed on social networks through CivilNet’s social media specialists. More information on the specialists mentioned above can be found at this link.
The #CivilNetCheck team also supports CivilNet journalists in preparing articles with accurate facts and without manipulation. If necessary, the team checks the facts in the other journalists’ materials, provides information from various open sources and registries and contributes to making the most accurate and fact-based CivilNet products.
Funding
CivilNet cooperates with local, international and Armenian Diaspora organizations and individuals within the framework of various programs. Information about them is available on the CivilNet website. The information about CivilNet and #CivilNetCheck’s donors and their logos are openly and fully posted to the CivilNet website.
Editors and Journalists Conduct
Political neutrality
We do not directly or indirectly support any political party or power, nor do we take sides in national or international conflicts or disputes.
For us, political neutrality and impartiality are fundamental principles. In order to avoid political conflicts of interest, CivilNet does not allow its journalists to engage in political activities or join or openly support any political power or party.
#CivilNetCheck must maintain absolute impartiality during its activities and not express any political preferences. Furthermore, funding or support by any political party or power is prohibited.
Journalists’ obligations
Members of #CivilNetCheck are obliged to be honest, impartial in their materials, refrain from expressing personal opinions and attitudes, be transparent about their sources and refrain from selectively choosing facts.
When checking claims and statements attributed to politicians and public figures, journalists are obliged to make sure that the statement was made by the person. In addition, journalists are obliged not to manipulate or distort quoted speech and to present it in the correct and adequate context.
We never assume, and we make sure and verify all the facts and data.
We use and accept only official documents as a basis for reporting.
We do not rely on the facts published by other media. No matter how reliable they seem, we always double-check them.
We are as accurate as possible when working with data and numbers.
We avoid evaluative judgments that are impossible to prove.
#CivilNetCheck encourages team members to regularly attend various courses and conferences to gain new knowledge, develop new skills and learn new tools. For maximum credibility, our journalists continuously strive to improve their skills in ensuring accuracy, which is one of the fundamental principles of the organization’s mission.
How do we choose which claims to check?
On a daily basis, our journalists are looking for statements and claims “worthy” of being checked. The team monitors, watches and reads statements, interviews, speeches, statements and messages from politicians, public figures, political powers, state and government bodies, as well as their posts on social networks. They monitor broadcast television, online and print media and social media.
Regional (Russian, Iranian, Azerbaijani, Georgian, Turkish) and international media publications and news related to Armenia and Armenians are also at the center of the journalists’ attention. For #CivilNetCheck, it doesn’t matter which political power or figure makes the wrong claim or statement. The team addresses all political powers.
Thus, when we notice statements, facts, or claims that seem suspicious and may be inaccurate, misleading, or delusive, we identify and check them.
First and foremost, we try to contact the person or organization that made the claim, asking them to provide the facts and data that prove it. If a person or organization provides facts to support their claims, and we are convinced that they are true, we do not address them. Exceptions are the cases when there is also doubt and discussion among the public regarding the accuracy of that statement or data.
However, if the person or organization does not cooperate with us, does not provide facts to prove their claims and statements, or the data provided does not prove the truth of their claims, we check those claims ourselves.
Since we cannot verify all claims, we select the most relevant and important ones
Our selection of fact-checked materials is broad and based on the following criteria:
Does the statement seem wrong or misleading?
Is the topic relevant and of public interest?
If it is wrong, can it harm the public or democracy by distorting public opinion? We do not address slips of the tongue and minor mistakes.
Are there facts in the statement that are verifiable and can they be singled out? We never verify or comment on opinions, predictions and rhetoric.
After reading or listening to the statement, will an average person have questions as to its accuracy?
How much has the information spread? We investigate whether the statement or claim has been shared and reprinted by media outlets with large audiences and high visibility, as well as by various individuals and platforms on social media.
Is it possible for the information to be further disseminated or repeated by other people? We also study how quickly information spreads and generates interactions.
As stated, we do not deal with opinions and predictions. The statement that is to be checked must contain certain facts or information, the accuracy or reliability of which can be verified. There are many examples of factual statements, but here are the most common:
The statement being checked must contain specific facts or information whose accuracy or credibility can be verified. There are many examples of fact-based statements; the most common include:
historical data,
comparisons,
legal aspect,
statistics
Another thing to remember: verifying the statement must be feasible, within a reasonable time and with the resources available to us. Naturally, there will be announcements that we would like to check, but currently do not have the access or resources to do so.
Media
As #CivilNetCheck’s mission is to reduce the spread of lies, fake news and manipulations and to fight misinformation, from time to time we have to address the media and the fake news they spread.
The golden rule for #CivilNetCheck is that if the fake news has not received much attention, we never write about it ourselves, so as not to contribute to the spread of fake news. The team’s focus is on issues and topics that are important to the public.
Social networks
As social networks have become popular among the public in recent years and are a major source of information for many people, #CivilNetCheck monitors social media as well, trying to identify false posts and materials that are going viral. The team also pays attention to posts and publications that have gained wide distribution and circulation on social networks, becoming the subject of public discussion.
Readers as #CivilNetCheck’s employers
The success and effectiveness of #CivilNetCheck is also directly related to reader’s engagement and active participation. Because the news cycle is so large and saturated, we may miss false and misleading publications, materials and posts, so it is important for us that readers send us such publications for checking, should they notice such posts – and we encourage readers to do so. In addition, readers can submit claims, materials, posts and questions they find suspicious, through the CivilNet website or Facebook page, and we will make every effort to check them.
We are convinced that a united and unified fight against misinformation and lies is essential.
For #CivilNetCheck, it’s also important to help hold the government, authorities, government agencies, officials and political forces accountable to the public. All these persons and institutions regularly make promises.
We collect both the promises made during the pre-election campaign, as well as study the action plans of the government and communities, follow the statements of politicians, ministers and other officials, their interviews, press conferences, and notes.
“Following up” on promises made
#CivilNetCheck addresses only promises that are measurable and verifiable.
Through our promise-meter, we evaluate whether a given promise has been fulfilled or not, or at what stage it is. It has four categories.
Completed: This rating is given when the initial promise has been mostly or completely fulfilled.
In process. This rating shows that steps are being taken to fulfill the promise.
Not fulfilled. A promise is categorized this way if it has not been fulfilled or the steps taken to fulfill it are insignificant.
Uncertain. When it is impossible to find any information about the steps taken to fulfill the promise.
The fact checking process
The fact-checking process and tools may differ in each case. They may include a thorough search on Google and other search engines, researching online open sources, registries, surveys, calls, using various tools (especially for images and videos), and consulting with various experts.
#CivilNetCheck relies on primary sources and original documents. It directly accesses government reports, academic studies, and other data. It is not enough for us to get data and information second-hand.
When an investigation reveals that a claim or statement is untrue, false, manipulative, out of context, or false and meets our standards for fact-checked material, #CivilNetCheck addresses that claim in an article, video, infographic, or Facebook post. The choice of format is made as a joint decision between editors and journalists.
Since transparent, unbiased and responsible journalism is the main principle of #CivilNetCheck’s activity, we present the sources in our materials, attach links to all sources of analysis, and, if necessary, provide answers to inquiries and recordings. The goal is for the reader to also check and make sure that the investigation and the data used are accurate, if they wish to do so. #CivilNetCheck does not use sources that cannot be identified.
Corrections and Complaints Policy
Transparency is important in both receiving information and presenting it. Although the #CivilNetCheck team makes every effort to ensure the accuracy of the materials, errors cannot be avoided. The team fixes them quickly and clearly. They do not hide and do their best to make corrections appear in the same news cycle, highlighting the fact that a correction has been made.
Readers and others can contact us by email, via our Facebook page or by calling us directly in case they spot a mistake in our articles or to dispute any material. #CivilNetCheck will review all submissions to ensure they are correct, and will correct any errors in the published materials.
If a fact-checked material contains an essential, factual error, we make a correction. An alert text is placed at the top of the article explaining what has been fixed and why. Then, the correction is made in the text of the article and, if necessary, is reflected in the article’s title and verdict. It is also possible to refer to and explain the same topic in a separate article.
As a member of the International Fact-Checking Network (IFCN) and the European Fact-Checking Standards Network (EFCSN), #CivilNetCheck adheres to the principles and standards established by both organizations.
If you believe that we have violated these principles, you may report your concern to either the IFCN or the EFCSN using the links below.
#CivilNetCheck Corrections
https://civilnet.am/hy/news/1009089
2024
https://civilnet.am/hy/news/781922
If you suspect a violation, please report it to IFCN via this link.
The Article’s Verdicts
#CivilNetCheck evaluates claims and statements with a number of ratings. We have 8 ratings by which we evaluate claims and statements.
Ratings of the Truth-meter and their definitions.
Correct. The statement is completely accurate.
True&false. The statement contains both true and false elements
Mostly false. The statement contains elements of truth, but key facts that would create a different impression are missing.
Incorrect. The statement is completely incorrect.
Misleading. Information that has been taken out of context, falsified, or altered, with a distortion of real information and misleading arguments based on correct facts.
Groundless. The statement is not substantiated; there are no available facts and data to support it.
Contradiction. The claim contradicts earlier statements.
Fake. Completely fake content.
Only Facts
Keeping in mind that misinformation and false information are often widespread because there are no factual and accurate materials and references on these topics, #CivilNetCheck also prepares factual articles that provide the most accurate information from reliable sources on a given issue.
To ensure the editorial independence of our work when dealing with funding from local, foreign or international sources, we publicly list funding sources for each project, including the projects that fund #CivilnetCheck section, on our web page. When dealing with the donors our team is not constrained in any way to address the various
DontGetScammed
#CivilNetCheck presents common scams, fraudulent schemes, and manipulative tactics aimed at deceiving people and causing financial and personal harm.
Meta verdicts
Meta prioritizes the verification of false claims that are timely and can have a widespread impact. The program does not cover opinions, obvious satire, humorous content, entertainment-related gossip, or business disputes. Additionally, the speeches and opinions of politicians, including their posts and advertisements, are not subject to verification.
Meta defines politicians as candidates running for office, current office holders, and, by extension, many of their senior appointees, like as heads of government agencies, along with political parties and their leaders. However, the posts of former officials and former candidates in elections are subject to verification according to Meta’s standards.
Therefore, fact-checkers on Facebook and Instagram review and assess posts, including advertisements, articles, photos, videos, Reels, audio recordings, and apply the appropriate ratings if they detect incorrect, misleading, or distorted facts or content.
Meta defines the following ratings:
FALSE. Content that has no basis in fact. This includes:
Fake quotes.
Claims that are impossible, or that could not be considered an interpretation of something that happened or was said.
Example: Claim that a natural disaster took place, when no such event happened.
Example: Claim that an individual created or patented something when they didn’t.
Conspiracy theories that explain events as the secret work of individuals or groups, which may cite true or unverifiable information but present an implausible conclusion.
Example: Claim a company is secretly engaged in drug trafficking based on an unrelated issue of charging high prices.
Example: Claim without evidence that government insiders are directly responsible for a terror attack to provide a pretext for going to war.
Fabricated content from websites misrepresenting themselves as real news outlets.
Image, audio or video content that is authentic but offered as proof of a separate event (that is, false context).
Example: Authentic photo that claims to show no damage in a town after an incident, but was taken before the incident.
Example: Authentic video that claims to show one person confessing to a crime, but is of another person.
Example: Presenting an authentic but old government ordinance as if it were current, when a new ordinance contradicts the old.
ALTERED. Image, audio, or video content that has been edited or synthesized beyond adjustments for clarity or quality, in ways that could mislead people about something that has no basis in fact. This includes media that is digitally created or edited through the use of artificial intelligence (AI). It also includes splicing media together, but not media excerpts (see ‘missing context’ rating), presenting media in a false context (see ‘false’ rating), or media where the digital creation or editing method is not itself misleading but the media includes a false claim (see ‘false’ rating). This does not include media with transparency using one of Meta’s AI labels or Meta’s AI watermarks or non-Meta transparency methods that clearly convey the use of AI, though a post may be eligible for another rating if it includes a claim separate from the use of digitally created or edited media. This definition includes:
Manipulated or transformed audio, video, or photos.
Example: Changing the speed of a video to misleadingly alter the speech qualities of the speaker.
Example: Adding an image into an authentic photo to present the appearance of something that never happened.
Media edited to omit or reorder the words someone said to reverse the meaning of the statement
Example: Removing the word “not” from someone saying “I will not do X.”
Synthetic image, audio, or video, including media created using artificial intelligence:
Example: A photorealistic image depicting an event that did not occur.
Example: A video depicting someone saying or doing something they did not say or do.
3. PARTLY FALSE. Content has some factual inaccuracies. This includes:
Inaccuracies or miscalculations regarding numbers, dates, times, but that could be considered an interpretation of something that happened or was said.
Example: Misstating the number of people registered for or attending an event.
Example: Miscalculating the cost of a government program.
A mix of true and false key claims, where the false claims do not predominate
Example: A list of several claims, some that are true and some that are false.
Example: A video that contains many key claims, some that are true and some that are false.
Content presented as an opinion but based on underlying false information
Example: Advocating for a policy change supported by several key claims, one of which is provably false.
4. MISSING CONTEXT. Content that implies a false claim without directly stating it. This includes:
Clips or excerpts from authentic media that hasn’t been altered (see “altered rating” definition) or presented in a false context (see “false rating” definition), but distorts the meaning of the original content to imply a false claim. This does not include clips of authentic politician speech (example, text quotes, video clip) (see Politician definition).
Example: An unaltered video clip of a group chanting that they’re peacefully protesting, but the full version of the video shows the same group instigating violence.
Example: Pairing together authentic text excerpts of someone’s speech that changes but doesn’t reverse the meaning of what they said.
Reporting on a false claim made by a third-party without questioning the veracity of the claim. Learn more about our guidelines for reporting false claims that were made elsewhere.
Example: A TV host airs an interview with a source who makes a provably false assertion, and the host doesn’t affirm or question the veracity of the claim.
Use of data or statistics that implies a false conclusion.
Example: Selectively using data from a study to imply a conclusion that wasn’t made by the study.
Example: Highlighting correlated data to imply a causal effect that is verifiably false.
SATIRE. Content that uses irony, exaggeration, or absurdity for criticism or awareness, particularly in the context of political, religious, or social issues, but that a reasonable user would not immediately understand to be satirical. This may be from sites not clearly labeled as or widely known as satire, or presented without clear labeling. Content rated as “Satire” will include fact-checkers’ articles for more context.
TRUE. Content that contains no inaccurate or misleading information.
What actions are taken when a post is rated according to META’s guidelines?
When #CivilNetCheck identifies a post or video containing false or misleading information, we first write an article detailing why the claim is incorrect, attaching all evidence and sources used so that readers can also verify the claim themselves.
We then enter the article into Meta’s system, where the appropriate rating is assigned. Based on this, Meta applies certain restrictions.
A label appears on the post, indicating that the claim, video, or photo is “False,” “Altered,” “Partly False,” or “Missing Context,” providing an option to learn why it was rated as such. The visibility of posts with such labels is reduced on Facebook, Instagram, and Threads.
If an account repeatedly posts content containing misinformation, Meta imposes restrictions not only on the post but also on the user’s account, page, group, or website. These restrictions last for 90 days. The visibility and reach of posts from these accounts are reduced.
Additionally, other limitations may be applied, such as restrictions on monetization and advertising opportunities, or a ban on creating new pages.
Meta does not remove false or misleading posts. Instead, it applies labels to them and reduces their visibility. Meta may remove posts only in exceptional cases that violate its Community Standards, such as misinformation or unverified news that could pose immediate risks of violence or physical harm. More details can be found at this link.
What to do when a post receives a label?
When Meta applies a label to a post, the author receives a notification. To remove the label and corresponding restrictions, a user can either correct the false content or appeal the decision.
Corrections: To issue a correction, you should correct the false content. For stories, add a subsequent story with the correct info saying you made a correction. If you need to fix a URL, make sure to correct the relevant information on your website and the Facebook or Threads post. For corrections to image, video, or audio posts, update the caption to correct the false content and clearly state that a correction was made. You can also link to an additional post that includes an updated, accurate version of the image, video, or audio, or to a fact-check article.
Disputes: To dispute a rating, you must clearly indicate why the original rating was inaccurate. Include a link to a source that supports your explanation for why a rating is inaccurate.
Users can send correction or appeal requests either to [email protected] or through Meta within seven days of receiving a notification. Fact-checkers will review those requests within three days and inform users of the appropriate final decision.
One more important note: Do not delete a post once it has already received a label, as this will not help in removing the associated restrictions.