The AI-powered fact checker investigating QAnon influencers shares its secret weapon

0
172

When it comes to QAnon, there’s a key player behind most of the research into who is responsible for some of the movement’s far-right conspiracy theories. Fact-checking tech company Logically has the failed screenwriter who became the big QAnon promoter, the Citigroup manager who secretly runs a big QAnon hub, and the QAnon platforms who worked with 8kun owner Jim Watkins, among others connected are revealed.

Founded in 2017 by CEO Lyric Jain, the company combines its AI technology with human fact-checking experts to uncover, track, archive, and debunk conspiracy theories ranging from QAnon to misinformation about COVID-19. The company already has some publicly available products like mobile apps and web browser extensions that allow users to navigate online misinformation they may encounter during their daily internet usage.

Now the company has announced that it will pass on to third parties one of the secret weapons with which QAnon is used against QAnon: Logically Intelligence, an AI-supported web platform that finds and analyzes online misinformation and offers countermeasures.

“Logically, intelligence is the culmination of four years of work,” Jain told me. “We can find ways to predict the impact of emerging threats and give our partners ways to mitigate them.”

While the platform is currently only available to a select few government agencies around the world, Logically Mashable offered an exclusive online tour of Logically Intelligence last week. Here’s what it can do.

Enter the situation room

Logically Intelligence enables users to keep track of multiple topics or narratives online, such as: B. Certain keywords or types of conversation around a conspiracy theory. Each conspiracy topic or misinformation topic is divided into individual “situation rooms”. For example, a user who wants to track COVID-19 misinformation can set up a situation room around that narrative, and Logically Intelligence tracks online conversations about these coronavirus conspiracy theories and disinformation campaigns across the web.

A QAnon Situation Room in Logically Intelligence provides a dashboard with a brief summary of the monitoring.

The screenshot above shows a situation room monitoring the QAnon content. As you can see, most of the Logically Intelligence content found related to the conspiracy still comes from these two platforms, even though Facebook and Twitter have banned QAnon from its platforms.

Logically Intelligence can pull content from social media platforms like Facebook, Twitter, YouTube and Reddit, as well as news articles and websites from across the Internet. Anyone who has ever used a social media or press monitoring tool like Sprout or Awario is very familiar with this part of Logically Intelligence.

Scan even the dark corners of the web

One of the things Logically Intelligence does that these other surveillance platforms don’t is track content from emerging alternative social media services. Logically’s platform can track content from sites like Parler, Gab, and even public group chats on Telegram.

This is an extremely important feature as many misinformation distributors and conspiracy theorists, including QAnon influencers, are moving to these alternative platforms. While mainstream social media sites can tell users where and how falsehoods are spreading, the source of these claims often lies on these alternative platforms.

Logically, Mashable says it continues to be preoccupied with how its customers use the platform. In fact, certain features – such as the ability to track users and content on Parler, Gab and Telegram – are only activated upon request to trusted partners. (Logically, I have to keep updating his platform to get this data as social media services are constantly changing their technology, which is why it is not currently enabled as the default.)

How it works

COVID-19 misinformation threats can be analyzed in detail by the platform.

Individual content is broken down by the AI ​​from Logically Intelligence to provide numerous data points.

Take the above screenshot of an anti-Vaxxer post pulled from the platform. Logically Intelligence analyzes the content to provide details on how far a particular account and the audience the post is targeting, as well as the audience. The AI ​​is also able to uncover recurring or new narratives from these contributions. Based on this information, Logically Intelligence categorizes misinformation into threat categories such as false information, toxic information, and inauthentic accounts.

An extremely important feature here is that every post retrieved from Logically Intelligence is archived directly on the platform so that it can be accessed regardless of whether the original poster or the social media platform removes the content.

Investigate how misinformation spreads to Logically Intelligence.

Investigate how misinformation spreads to Logically Intelligence.

Logically, intelligence also provides a visual experience to explore the data. The screenshot above almost looks like a beautiful piece of art … until you remember that it ties together some of the most bizarre and even dangerous conspiracy theories and nefarious actors on the internet. This web cloud helps users connect the dots between various misinformation. Users can see which accounts post certain hashtags or articles and how they are spread across the internet. All of this is extremely useful for users connecting threads or wanting to find the source of falsehoods, and for monitoring certain influential users and narratives that are being caught.

Find the bots and prevent bad actors

The platform doesn’t just give out misinformation. It can also expose accounts armed for conducting disinformation campaigns like bot accounts, accounts impersonating other users, accounts that use automated behavior, and even extremist content.

Taking countermeasures, Logically Intelligence users can take action against the threats posed by misinformation they discover through the platform.

Taking countermeasures, Logically Intelligence users can take action against the threats posed by misinformation they discover through the platform.

This is where the countermeasure function comes into play, with the help of which actual action can be taken against a user or content. Logically Intelligence enables users to require human misinformation experts to look at a specific claim to verify key facts. You can also report the content to social media companies like Facebook and Twitter directly on the platform.

Logically Intelligence users can request misinformation investigations conducted by human fact checkers.

Logically Intelligence users can request misinformation investigations conducted by human fact checkers.

Currently, the platform is only being tested with selected government agencies around the world. While Logically did not disclose exactly who is using the platform, the company said that Logically Intelligence was used during a 2020 election in a U.S. battlefield state to identify 40,000 threats to electoral integrity. (Logically, the platform has discovered more than 3.4 million high-threat content in the past 12 months.)

A core user base made up of government agencies has raised numerous concerns for me. For example, what if a government agency tried to use the platform to classify critical factual information about the ruling administration as misinformation? Examples range from Myanmar’s new government cracking down on coup plotters to US law enforcement agencies monitoring Black Lives Matter activists and promoting anti-protests last summer.

According to the CEO of Logically, the company monitors how customers use Logically Intelligence. This is not difficult, at least for the time being, as only a few selected third parties use the service. Human fact checkers analyze the veracity of the claims on the platform so not everything depends on the AI, which may miss the nuance in some of the above scenarios. In addition, any use of the platform that deviates from its purpose of monitoring misinformation would need to be approved by the company’s ethics committee. Logically, it is selective with who it works with and contractually reserves the right to end a partnership at any time.

Facing technical challenges and beyond

The biggest challenge for Logically Intelligence isn’t just technological. The company needs to keep working on its AI to keep it updated with the ever-changing landscape of the internet. Logically, the intelligence will also require its human fact checkers to keep their finger on the pulse of online dissemination. As Logically Intelligence expands its user base from just a few select government agencies, the company also needs to ensure that its platform is not being used by bad actors.

Anyone who uses the platform in good faith must also know that this tool does not solve the problem of misinformation. However, by using Logically Intelligence, one can definitely help keep track of the problem, better understand its spread, and suppress certain false narratives. They may even expose some of their own QAnon influencers.