Bust-AD: Russian-Backed Deepfake Targets Lithuanian News Reporter

The fact-checking initiative Debunk.org reported that a Russian-linked deepfake video featuring a Lithuanian-language news presenter from the 3 Plus channel was used in fraudulent Facebook ads claiming to offer victims of online scams a chance to get money back through … Read More

Press release from Debunk.org

Debunk.org has released a groundbreaking investigation that exposes the inner workings of the so-called “Sprinter Network” — a disinformation ecosystem operating across Europe through a series of websites, social media channels, and pseudo-media platforms. This network, while designed to resemble … Read More

Typographic Attacks in a Multi-Image Setting

Typographic Attacks in a Multi-Image Setting Large Vision-Language Models (LVLMs) are susceptible to typographic attacks, which are misclassifications caused by an attack text that is added to an image. In this paper, we introduce a multi-image setting for studying typographic … Read More

Fine-grained Fallacy Detection with Human Label Variation

Fine-grained Fallacy Detection with Human Label Variation We introduce FAINA, the first dataset for fallacy detection that embraces multiple plausible answers and natural disagreement. FAINA includes over 11K span-level annotations with overlaps across 20 fallacy types on social media posts … Read More

LLMs vs Established Text Augmentation Techniques for Classification

LLMs vs Established Text Augmentation Techniques for Classification: When do the Benefits Outweight the Costs? The generative large language models (LLMs) are increasingly being used for data augmentation tasks, where text samples are LLM-paraphrased and then used for classifier fine-tuning. … Read More

AI-CODE empowers media professionals to harness AI for good

Artificial intelligence holds immense potential to support journalism, fact-checking, and strategic communication — but only if used ethically and transparently. This video showcases the work of the AI-CODE project, which empowers media professionals to explore and adopt AI tools in … Read More

AI in journalism: The good, the bad, the ugly

Author: Xhoi Zajmi This article is part of our special report AI-CODE – Empowering Media Professionals against Disinformation. The rise and integration of AI in our daily lives is not a futuristic fantasy, and of course, journalism is not exempt. The … Read More

Graphic shows 5 circles and dots for a single account. Most of the dots are in the non-hateful area. Some are outside the circles. Racism and Sexism have most of the dots.

Between constructive dialogue and rage: Visualizing Bluesky posts with a new experimental data analyser

The Bluesky Data Analyzer Authors AI-CODE partner: Bahia Albrecht, Andy Giefer, Kay Macquarrie – Deutsche Welle Hateful speech is labeled into four categories: Racism, Religious, Sexism and Sexual Orientation. In addition, non-hateful and neutral posts are represented. The example shows … Read More

Transparency essential for developing trustworthy AI, says Gürtler

As Europe pursues innovation and competitiveness in AI, should AI be structured around human rights and sustainability; and does AI’s employment impact matter? Paula Gürtler deploys a moral compass. As artificial intelligence (AI) reshapes society, ethical questions are increasingly raised … Read More

Europe steps-up AI investment to counter fierce US-China competition

‘Game On’ for Europe? A €100 bn injection into France’s AI sector, the EU’s €200 bn for InvestAI, and €150 bn in venture capital for the EU AI Champions Initiative. The global AI race has reached a critical juncture, with … Read More