Co-Inform · Project

AI Platform That Detects Misinformation and Tracks How It Spreads Online

digitalTestedTRL 5

Imagine you could have a smart assistant that watches social media and flags false or misleading posts before they go viral. That's essentially what Co-Inform built — a platform with tools that spot misinformation, predict how it will spread through online networks, and even help break into echo chambers with corrective information. It was designed with three groups in mind: regular people who want to know what's real, journalists who need to verify claims fast, and government officials who make decisions based on public sentiment. Think of it as a fact-checking radar combined with a rumor-tracking map.

By the numbers

consortium partners involved in development

European countries where platform was tested

architecture versions iteratively refined

total deliverables produced

demonstration deliverables with working prototypes

SMEs involved in the consortium

The business problem

What needed solving

Misinformation spreads faster than facts on social media, damaging brand reputations, distorting public policy, and eroding trust in legitimate news. Companies and governments currently lack automated tools that can detect misleading content early, predict how it will spread, and intervene before it reaches critical mass. Manual fact-checking simply cannot keep up with the volume and speed of online misinformation.

The solution

What was built

The project built the Co-Inform platform — a modular system with content collection services and APIs for retrieving and semantically enriching social media content, services for detecting and categorizing misinformation, tools for measuring and tracking public perception and behavior toward false claims, and a user-facing interface. The architecture was refined through 4 documented versions with interoperability at business and information level as a design priority.

Audience

Who needs this

Digital news publishers needing automated fact-checking at scalePR and brand reputation management firms tracking misinformation threatsGovernment communications teams monitoring public sentimentSocial media platforms building content moderation toolsElection integrity organizations tracking disinformation campaigns

Business applications

Who can put this to work

Media and Publishing

any

Target: Digital news publishers and media monitoring firms

If you are a news publisher or media monitoring company dealing with the constant challenge of verifying user-generated content — this project developed a platform with content collection services and misinformation detection APIs that can flag misleading posts across social media. The Co-Inform architecture went through 4 documented iterations with 12 partners across 7 countries testing it, giving it real multilingual coverage.

Corporate Communications and Brand Safety

mid-size

Target: PR agencies and brand reputation management firms

If you are a PR or communications firm struggling to protect clients from viral misinformation campaigns — this project built services for measuring and tracking public perception and behavior toward misinformation. The platform can predict which misleading content is likely to spread across specific demographic sectors, letting you act before a false narrative damages your client's brand.

Government and Public Policy

enterprise

Target: Government agencies and policy research institutes

If you are a government agency or policy institute that needs to understand how misinformation affects public opinion on health, economy, or environment — this project created advanced misinformation analysis tools specifically designed to support evidence-based policy making. The platform was co-created with policymakers and tested across 7 European countries, making it ready for cross-border deployment.

Frequently asked

Quick answers

What would it cost to deploy this misinformation detection platform?

The Co-Inform tools and platform were designed to be freely available and open sourced, which means the software itself has no licensing fee. Your costs would be infrastructure (cloud hosting, API usage) and integration effort. A 12-partner consortium built this over 3+ years, so the underlying R&D investment was substantial.

Can this scale to monitor millions of social media posts in real time?

The platform includes content collection services with API access for automated content retrieval and semantic enrichment. The architecture went through 4 documented versions, each refining scalability. However, real-time processing at enterprise scale would likely require additional infrastructure investment beyond the research prototype.

What is the IP situation — can my company use this commercially?

The project explicitly stated that tools and platform will be made freely available and open sourced to maximize benefit and reuse. This is favorable for commercial adoption, though you should verify specific license terms on the project outputs. As an EU-funded RIA project, results are typically accessible under open licenses.

Does this work for non-English content and markets?

The consortium spans 7 countries (Austria, Cyprus, Germany, Greece, Spain, Sweden, UK) with partners likely contributing multilingual capabilities. The content collection services deliverable specifically mentions language coverage as a documented feature. Check the API documentation for currently supported languages.

How accurate is the misinformation detection?

Based on available project data, the platform includes services for categorizing behavior toward misinformation and measuring opinion shifts. The project produced 26 deliverables over 40 months of research, including dedicated demonstrators for perception tracking. Specific accuracy benchmarks would need to be verified in the technical deliverable reports.

What is the timeline to integrate this into our existing systems?

The project delivered a documented API for content access and a generic architecture designed for interoperability at business and information level. With 4 architecture versions refined over the project lifetime and open-source availability, integration timelines depend on your existing tech stack, but the API-first design should reduce friction.

Consortium

Who built it

The Co-Inform consortium brings together 12 partners from 7 countries, with a mix of 6 universities, 3 industry players, 2 research organizations, and 1 other entity. The 25% industry ratio and 4 SMEs suggest meaningful private-sector involvement, though the project was university-led (coordinated by Stockholm University). For a business looking to adopt these tools, the academic backbone means strong research rigor but may require additional engineering to reach production-grade quality. The geographic spread across Austria, Cyprus, Germany, Greece, Spain, Sweden, and the UK provides good European coverage for multilingual misinformation challenges.

STOCKHOLMS UNIVERSITETCoordinator · SE
TECHNOLOGIKO PANEPISTIMIO KYPROUparticipant · CY
SCYTL SECURE ELECTRONIC VOTING SAparticipant · ES
SCYTL ELECTION TECHNOLOGIES SLparticipant · ES
DIETHNES PANEPISTIMIO ELLADOSparticipant · EL
INTERNATIONALES INSTITUT FUER ANGEWANDTE SYSTEMANALYSEparticipant · AT
UNIVERSITY OF STUTTGARTparticipant · DE
UNIVERSITAT KOBLENZparticipant · DE
THE OPEN UNIVERSITYparticipant · UK
EXPERT SYSTEM IBERIA SLparticipant · ES

How to reach the team

Stockholm University, Sweden — reach out to the Department of Computer and Systems Sciences for project leads

Order a report on this consortium See STOCKHOLMS UNIVERSITET profile →

Next steps

Talk to the team behind this work.

Want to deploy misinformation detection for your organization? SciTransfer can connect you directly with the Co-Inform research team and help evaluate fit for your use case.

Order a report on this topic More in Digital & ICT