ChatGPT powers automated content, spam farms
Content farms powered by ChatGPT and similar chatbots are spewing low-quality posts and spam in multiple languages, according to a report by NewsGuard. The researchers discovered 49 examples of 'news' sites apparently intended to draw clicks and attract advertising revenue covering politics, technology, finance and celebrity news in Chinese, Czech, English, French, Portuguese, Tagalog and Thai . NewsGuard says that whilst some content is demonstrably false (' Biden dead . Harris acting President, address 9am ET.') , most people would not be able to tell if the content is generated by AI as it is not labeled as such. Furthermore, much of it is riddled with errors, including error messages such as 'I can not complete the prompt', and is mostly unchecked or edited by human hand. Earlier, Vice News had reported that Reddit moderators are already experiencing a big increase in fake accounts and posts apparently generated by generative AI products like ChatGPT. System 🤖 ChatGPT Operator: OpenAI Developer: OpenAI Country: Brazil; China; Czechia; France; Philippines; Portugal; Thailand; USA; UK Sector: Media/entertainment/sports/arts Purpose: Provide information, communicate Technology: Chatbot; Generative AI; Machine learning Issue: Business model; Mis/disinformation
- Date it happened
- 2023-05-01
- Organisation involved
- OpenAI
- Product, system or model
- ChatGPT
This incident was imported from AIAAIC and is used under CC BY-SA 4.0. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.
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.