Study: Large language models can mimic QAnon

Large language systems such as OpenAI's GPT-3 can be used to create and deploy convincing fake news on social media, ac cording to researchers, prompting concerns about their use for misinformation and disinformation. R esearchers at Georgetown University's Center for Security and Emerging Technology (CSET) conducted a study to determine if AI models could be prompted to generate content indistinguishable from human-written conspiracy theories. By fine-tuning or specifically prompting LLMs, the study found that these models could effectively "mimic" the linguistic style, thematic elements, and logical fallacies of the QAnon movement, highlighting a transition from manual "trolling" to automated, high-volume radicalisation efforts. The potential impacts include the flooding of social media platforms with synthetic extremist propaganda, which can overwhelm content moderation systems and accelerate the radicalisation of vulnerable individuals. The researchers went on to argue it will become increasingly difficult to distinguish reliable and fake news and information. System 🤖 GPT-3 Developer: OpenAI Country: Global Sector: Multiple Purpose: Generate text Technology: Large language model (LLM); NLP/text analysis; Neural network; Deep learning; Mach ine learning Issue: Mis/disinformation; Safety

Date it happened

2021-05-01

Organisation involved

OpenAI

Product, system or model

GPT-3

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.