💬 We're here to help!

The teams at Logically and Logically Facts are here to help you make the most of this product, so if you have any questions, feedback or just want to say hi, you can email us at [email protected] 💚

You can also join our Slack Accelerate Community Workspace, where you can keep up to date with announcements, see tips and tricks, get help and report issues.

About Logically Accelerate

Characteristics of Multimodal misleading and deceptive media on Social Media

We know that campaigns of misleading and deceptive information can leverage digital infrastructure that is unparalleled in its reach. The internet reaches billions of individuals and enables bad actors to tailor persuasive messages to the specific psychological profiles of individual users. [1 ,2]

Misinformation has been identified as a contributor to various contentious events, such as, elections, it has been instrumental in political or religious persecution, and caused harm during the COVID-19 pandemic. [3, 4 ,5]

The current online media environment is increasingly multimodal – comprising of both visuals and text [6] . In fact, whilst fact checking the Lok Sabha elections our team in India found more than 50% of content that was fact checked, was video. This is partially impactful because research shows that people are likely to believe inaccurate visuals more than mere text [7].

Worryingly, recent progress in Generative AI has led to concerns that it will “trigger the next misinformation nightmare” [8], that people “will not be able to know what is true anymore” [9], and that we are facing a “tech-enabled Armageddon” [10].

There is more debate than ever around the truthfulness and accuracy of information [11].

Challenges For Fact Checking

Timely fact-checking is crucial in combating misinformation and minimizing its impact [12]. However, traditional methods often demand significant time, with approximately 60% of fact checkers' time consumed during initial claim discovery. [13]

Therefore we see a key challenge for fact-checking is that it takes hours of time to discover fact-check claims in videos, before even getting to the verification process.

The time to response problem

Our goal is to reduce the time that humans spend discovering content, with the aim of helping to produce a greater number of more timely fact checks, increasing the impact of the crucial work fact checking teams do.

Logically Accelerate uses AI to help fact checkers surface net new misinformation as quickly and painlessly as possible. We are focus on:

1. Proactively identifying and analysing multimodal content
2. Increasing visibility of harmful fact checkable claims
3. Accelerating content assessments

With the aim of optimising the time to response.

Logically Accelerate speeds up the time to response

By reducing the time spent discovering what might be fact checkable, Logicaically Accelerate can drastically reduce the time to response for any given claim.

When we spoke to fact checkers about their process we came up with the following four phases:

• CONFIGURE: where the user sets up the environment and identifies keywords for searching.
• DISCOVER: where the user is finding the relevant content from within the sea of information that is returned by the search.
• EVALUATE: where the user assesses each piece of content and chooses what is most urgent for further investigation.
• ACT: where the user decides to produce a fact check or not, by looking at the media and claims and making a judgement or assessment.

We know that a significant amount of time is spent in the discovery phase - and this is why Logically Accelerate focuses on the discovery phase of the process, and is designed to accelerate the discovery of harmful fact checkable content. Our aim is to give you back the time you spend looking for content, so you can invest that time more in the verification process.

[13] Measured time taken by Logically Facts Fact Checkers during Video Fact Checking in a European context.