Artificial Intelligence (AI) is changing and improving the cybersecurity industry, offering both enhanced defense mechanisms and automated security systems. However, as organizations work to strengthen their security posture using AI tools, threat actors are also leveraging these technologies to enhance their attacks. Recent findings from Google’s Threat Intelligence Group (GTIG) reveal how these actors are currently using AI tools like Gemini.
GTIG’s research indicates that government-backed hacker groups are using the Gemini to boost their productivity. While threat actors attempt to weaponize AI tools, their use of Gemini remains limited to research, coding assistance, and content creation. Google says that the threat actors are yet to develop new AI-based attack methods but use these tools to reduce preparation time.
Hacker groups from over 20 countries use Gemini primarily for:
- Conducting research to discover potential targets and free hosting providers.
- Gathering information about target organizations.
- Researching vulnerabilities in systems.
- Developing harmful software.
- Writing malicious scripts and finding ways to avoid detection.
Hackers also use Gemini for influence operations such as developing convincing false personas, creating target messages campaigns, and translating content for different audiences. The study showed that Iran and China were the most active users of Gemini, while Russian groups showed limited activity during the study period.
When hackers try to use Gemini for more harmful purposes, it responds with the system’s safety messages and refuses to help. Despite several attempts to bypass these protections using publicly available jailbreak techniques, their efforts have been unsuccessful.
These constant threats to the safe use of AI highlights the importance of collaboration of the private sector, governments, and other stakeholders to work together to ensure AI development remains beneficial while minimizing potential harm. Google continues to enhance its AI models and strengthen defensive capabilities through ongoing research and development.
About the author:
