AI Agent Carries Out First-Ever Unassisted Ransomware Attack

Autonomous AI Agent Deployed in JadePuffer Ransomware Incident

According to НВ — Техно: On July 7, 2026, cybersecurity firm Sysdig reported a landmark event: the first known case of an autonomous AI agent being used in a ransomware attack, dubbed JadePuffer. In this operation, a human operator selected the target, configured the command-and-control servers, set up data exfiltration storage, and supplied the agent with pre-obtained credentials. The AI agent then exploited a vulnerability in the open-source tool Langflow to carry out the attack.

Attack Details and Fallout

The agent infiltrated a MySQL server, escalated its privileges to administrator level, and encrypted over 1,300 configuration records. After an initial login failure, it found a workaround in just 31 seconds. To conclude the assault, the agent dropped a ransom note containing a Bitcoin wallet address. Stolen data included access keys for OpenAI, Anthropic, DeepSeek, and Gemini services.

'The operator chose the target, set up servers for attack management and stolen data storage, and handed over already obtained credentials to the agent.' - Michael Clark, Director of Cyber Threat Research at Sysdig

Sysdig has not disclosed the identity of the victim. Furthermore, the company was unable to determine which specific model powered JadePuffer. Microsoft researcher Jeff McDonald suggested the attack may have used an open-weight model with its safety mechanisms disabled.

Sysdig researchers noted that 'what impressed us most was not the sophistication of the methods, but the agent's speed.' They also warned that 'the use of such agents could allow cybercriminals to launch thousands of attacks simultaneously.'

This incident underscores the growing threat posed by autonomous AI agents in cybercrime. Their use in malicious activities could dramatically complicate defenses, as the speed and automation of attacks significantly increase risks for organizations. Experts are urging stronger security measures to protect critical data and systems from threats that may soon become commonplace.

The rise of autonomous AI in cybercrime is not an isolated incident. In a related report, experts discuss how AI technology is enabling novices to engage in hacking activities, creating a new level of cyber threats. Understanding these developments is crucial as they highlight the broader implications of AI’s role in facilitating malicious actions. For more insights on this alarming trend, read about how AI is empowering new hackers.