Recognizing Fake News. Technologies for Creating False Information

photo: Rawpixel.com/CC0/Pexels

Major companies have announced efforts to fight fake news, but the task of distinguishing real information from fake news is tackled daily by media monitoring companies.

- Sometimes it’s hard to tell if a comment comes from a fake account or a user who strongly believes in the truth of one side only, says Sebastian Bykowski, Vice President and General Manager of PRESS-SERVICE Monitoring Mediów, to Newseria Innowacje. - Tools for tracking history, monitoring, and profile analysis can give us significant information, helping us determine if an account is fake or a real participant in a discussion.

In the age of increasingly widespread fake news, bots publishing manipulated information, and internet trolls, finding reliable data online is challenging. Professional media monitoring companies, including those specializing in social media, use advanced tools to identify fake accounts and information. While the primary job of media monitoring companies is to find every possible piece of information in the media, clients increasingly want to know which information is real and which may be fake.


- Analysts select and examine the history of messages on individual accounts. It’s a relatively complex process, as it requires checking account history, the number of reports and posts over time, and whether they target a specific audience or follow a particular message line, says Sebastian Bykowski. - When we find that a particular source consistently produces negative, aggressive content, we alert our clients, identifying particularly active or significantly more active media outlets.

Artificial Intelligence Fakes Information

Artificial intelligence is increasingly used to fake information. There is now software that can manipulate public appearances by politicians. For example, a video circulates online showing Barack Obama giving the same speech at different ages – edited using specialized software. This means that anyone’s public appearance can be faked.

An example of large-scale use of fake information was the Facebook scandal in early 2018. Cambridge Analytica collected data from over 50 million Facebook users and used it to create targeted ads influencing specific voting decisions. This likely impacted the Brexit vote and the recent U.S. presidential election.

- In politics, there are specialized entities that create specific content and comment on it as widely as possible, but there’s a visible trend of more bots and various automated services appearing with the goal of spreading and copying negative content quickly and widely, the expert explains. - When we add micro-targeting tools and ads aimed at specific audiences with targeted messages, it all contributes to a highly intensified negative impact.

• Facebook CEO Mark Zuckerberg appeared before U.S. Senate committees to explain how user data was used and also testified before the European Parliament.
• As a result of the scandal, Facebook administrators deleted 583 million fake accounts and billions of posts violating community standards.
• Zuckerberg announced further steps to fight bots and internet trolls.

Bots are also increasingly used in Poland, especially in politics.

• In early August 2018, it was noticed that Patryk Jaki`s campaign (from PiS) was criticized on Twitter by users from Pakistan, Nepal, and Reunion Island.
• In late October and early November 2017, it was observed that Rafał Trzaskowski`s (from PO) account was followed by 10,000 inactive Twitter accounts identified as potential bots for the municipal campaign.

- Many recent studies, including those by Oxford professors, show that Poland is beginning to resemble other countries where aggressive, negative promotion of certain content is common, concludes Sebastian Bykowski. - It’s undoubtedly a very negative phenomenon. In today’s communication world, hate and trolling are among the biggest threats to the democratization of the internet and, by extension, public life.


According to a Transparency Market Research report, the media monitoring tools market will exceed $4.2 billion by 2022, with an annual growth rate of 13.6%.

Fake News in Poland and Europe

Three-quarters of Poles encounter fake news, i.e., false or distorted information, once a week or more often. Only 14% of respondents feel this problem does not affect them, according to the Kantar Public survey results from Eurobarometer.

To the question "Do news and information that distort reality or are even false represent a problem in your country?" 84 percent of Polish respondents answered affirmatively.

• 49% - definitely yes
• 35% - rather yes
• 9% - rather no
• 3% - definitely no
• 4% - don`t know/no opinion

In this respect, Poland ranks in the middle of the EU scale, with results comparable to the European average (44% - definitely yes, 41% - rather yes).

Eurobarometer survey results confirm the findings of a recent Press Club Polska analysis on trust in media information. It shows that we trust media most on social issues and least on political matters.

• social information - trust: 68%, distrust: 30%
• economic information - 57% trust, 40% distrust
• political information - 46% trust, 49% distrust

Context, neutral language, or even the popularity of the information source, according to the Press Club Polska survey, play a secondary role. The credibility of media information is most influenced by presenting opinions from independent experts, authorities in a given field, referencing scientific studies, and presenting multiple information sources.