Summary
Google security researchers found that hackers are using AI-based malware in real cyberattacks. They identified two types of malware, called PromptFlux and PromptSteal, that change behavior during an attack using AI technology. This shows that hackers are starting to use AI to make their attacks more powerful.
Key Facts
- Google researchers found AI-powered malware used in real-world cyberattacks.
- The malware strains, named PromptFlux and PromptSteal, use large language models.
- These malware programs can change their code to avoid being detected.
- PromptFlux was found while scanning a malware-checking tool called VirusTotal.
- PromptSteal has been used by Russian hackers against Ukraine.
- Both malware types are in early stages of development.
- Researchers are concerned about hackers using open-source AI models.
- The underground market for AI cyber tools has grown in the past year.
