4 AI chatbots run malware autonomously
The AI Brains Behind the Malware: How Attackers Are Building Autonomous Threats
The world of cybersecurity is constantly evolving, moving from reactive defense to anticipating the next move. Recently, researchers have uncovered a fascinating and worrying new dimension in how sophisticated malware operates: it is leveraging the very same artificial intelligence systems that we use daily to plan and execute its malicious actions.
The threat lies with a piece of malicious software known as CLOSEDQUORUM, which has been designed to operate with an unnerving degree of autonomy. Instead of following a simple, linear set of instructions, this malware doesn’t just execute code; it engages in complex, dynamic decision-making by consulting multiple large language models.
Specifically, CLOSEDQUORUM communicates with four distinct AI chatbots—DeepSeek, Qwen, Mistral, and Google Gemini—before determining its next move. This coordination is more than just a clever trick; it represents a significant leap in the sophistication of cyber threats.
What makes this finding particularly alarming is the resilience built into this system. The malware is engineered to maintain operation even if one of these AI services is temporarily unavailable. It achieves this by seamlessly querying the others, ensuring that the attack chain never halts, regardless of external service disruptions.
This level of redundancy transforms the malware from a single-point vulnerability into a highly resilient, self-healing threat. It demonstrates how attackers are utilizing the interconnectedness of advanced AI systems to create more autonomous and difficult-to-trace malicious entities.
For cybersecurity defenders, this discovery provides an invaluable early look into the future of threat modeling. It forces security professionals to rethink traditional defensive strategies, acknowledging that the lines between human-designed systems and autonomous digital agents are rapidly blurring.
Understanding this reliance on distributed AI coordination is crucial. It highlights the necessity of developing defenses that can monitor and disrupt these sophisticated, multi-agent attacks before they can cause widespread damage. The age of truly autonomous threats is here, and the defense must evolve as quickly as the attackers.
Buy on Amazon
Credit
