Summary
OpenAI released a report about foreign groups using AI tools, like ChatGPT, for hacking and influence operations. These groups use multiple AI models to plan and carry out their campaigns. OpenAI has banned several accounts linked to these activities, including ones tied to China and Russia.
Key Facts
- Foreign adversaries are using AI tools for hacking and influence efforts.
- ChatGPT is often used to help plan these operations.
- OpenAI banned accounts connected to state-sponsored campaigns from countries like Russia and China.
- A Russian-based actor used ChatGPT to create prompts for another AI model.
- Chinese-language accounts used ChatGPT to refine phishing tactics with another AI tool.
- OpenAI identified and banned accounts using AI for malware creation and phishing email writing.
- Investigators find it hard to see how specific AI models are used because threat actors use multiple tools.
- The effectiveness of these AI-driven campaigns is still limited.
