Researchers test MLLMs for generating and detecting fake multimodal social media posts
A team of researchers explored whether multimodal large language models (MLLMs) can create and spot fake news on social media. They developed a multi-agent system with a story agent, an image agent, and a critic agent to produce over 9,000 paired fake posts across science, health, and entertainment topics. The framework aimed to generate realistic counterfeit content that mimics true news posts.
Key points
- Researchers created a multi-agent system generating 9,000 fake multimodal social media posts in science, health, and entertainment
- Tested 16 MLLMs for detecting fake news, finding most models fall short of human-level accuracy
- Models performed poorly at identifying image authenticity in fake posts
The study benchmarked 16 open- and closed-source MLLMs for both generating and detecting fake multimodal posts. Results showed most models struggled to match human-level accuracy, particularly in identifying image authenticity. The findings highlight gaps in current AI defenses against disinformation and provide a foundation for future research. Code and datasets are available on GitHub.
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- Researchers release ArgGYM benchmark for testing defeasible reasoning in AI models · 1 src
- Researchers release SimTrace for generating synthetic user behavior data · 1 src
- MetaPersona framework uses 11,000+ studies to build synthetic populations for AI tasks · 1 src
- Researchers propose DLFP controller to cut AI inference latency by up to 30% · 1 src
- Researchers introduce GoldiMask to improve diffusion language model fine-tuning · 1 src
Comments
via GitHub Discussions