TOOLS ai_guments – AI-powered myth buster for social media work

The problem is not new, and there have already been various attempts to address it. Why another one? 

Many exciting projects have already been funded, but they have not been able to develop a viable business model from this funding. Other projects have been commercially successful, but are only affordable for very large organisations. Some focused on hate speech, which is difficult to define, or only on facts, without also capturing narratives. Many applications also function in isolation, are insufficiently integrated into the work processes of social media editorial teams and are therefore not very effective. Or they require third-party software for further integration.

We took a closer look at the following projects: 

Factinsect Funded AI project from Austria that, like Ai_gumets, used AI to verify statements. After the funding expired, the project was discontinued in 2025. The start-up was focused on providing fact checks for Meta, and with Mark Zuckerberg's announcement that he would discontinue this, the business case no longer applied. 
Differences: Unlike Ai_guments, the AI does not compare with an argumentation database but with a large amount of publicly available original sources, which, in our opinion, increases resource requirements and reduces accuracy. Users were only informed about the truthfulness of statements but were not supported in formulating responses, and there was no provision for storing their own counter-argument patterns and responses privately.

Factiverse is a Norwegian SaaS company that originally specialised in TV debates and can verify sources for claims as they are made. An application that can also be used to check new responses in ai_guments as they are being written, but also to identify false claims that ai_guments has overlooked so far. Factiverse offers essentially volume-based billing, which would be financially viable for pools of several thousand requests per month. We want to examine whether Factiverse can be used in Ai_guments as a source for identifying new narratives. 
Differences: The application is not originally specialised in social media, but in real-time fact-checking for live media. There are no plans (yet) to store our own counter-argument patterns privately

Gooogle Fact Check Tools Google offers a targeted search for third-party fact checks. Sources can be selected specifically for this purpose. We plan to offer some of these sources via the API in the ai_gumets data pool. 
Differences: Unlike ai_guments, however, users must first search for fact checks by entering a keyword, are not supported in formulating responses, and it is not possible to save their own counterargument patterns and responses privately.

Logically is a British company that claims to be able to monitor and classify the spread of disinformation across a wide range of channels, but only offers its services from £80,000 per year and works primarily for companies and intelligence agencies. Its focus is more on averting danger than participating in public debates. 
Differences: Presumably no support in formulating and disseminating corrections and counterarguments.

Mimikama is an organisation that has been exposing false reports and disinformation on social media since 2011, as well as exposing hoaxes and phishing methods. Mimikama is open to collaborations.
Differences: Unlike ai_guments, however, users must first search for a keyword after fact checks, are not supported in formulating responses, and it is not possible to save their own counterargument patterns and responses privately.

Pino Fact Checker is a Chrome extension that uses AI to check the factual content of selected text with a right mouse click, searches for sources and outputs a truthfulness score. Nice: The tool's mascot is a wooden doll whose nose grows as the truthfulness score decreases.
Differences: Ai_guments is designed to support various browsers and not only check facts but also provide counterarguments, not with text snippets but with individual responses tailored to the original post and context.