Testing Multimodal Generative AI: Generating Election Mis-and-Disinformation Evidence
Logically tested three image-based generative AI platforms to see if they would produce evidence of elections mis-and-disinformation narratives in the US, UK, and India, and found that more than 85% of prompts were accepted.
Report by: Kyle Walter
Study: Deanonymising Threat Actors
Understand how Logically uncovers those that leverage the anonymity of the internet to spread harmful misinformation and conspiracies using OSINT.
Climate Change Misinformation Report
APCO Worldwide and Logically undertook a collaborative research project focused on identifying and understanding the most prominent climate misinformation narratives ahead of COP26.
Video: Seeing is Deceiving - How Deepfakes Work
Find out how the relatively new phenomenon of deepfakes work.