Google’s Threat Intelligence Group reports on bad actors attempting to misuse its AI model, Gemini. Threats fall into two categories: using AI for tasks like malware coding and trying to manipulate AI into unsafe actions. Advanced Persistent Threats (APTs) from China, Iran, North Korea and Russia primarily use Gemini for research on military targets, software vulnerabilities and coding help. Information Operations (IO) groups use it for translation and disguising identities. While Google’s safeguards prevent direct harm, AI still aids bad actors in indirect ways. The full report provides deeper insights into these threats.
Source: Android Authority
