Spooky New Infostealer Malware Uses AI to ‘Autodox’ Its Victims

A newly discovered infostealer called Volta-Autodoxer uses large language models to build detailed profiles of infected computer users and help attackers identify valuable targets. Cisco Talos researchers found the AI features while examining the malware, though they have not confirmed whether it has been used in real-world attacks.

Infostealers are often disguised as counterfeit software. Once installed, they can harvest passwords, authentication tokens, browser cookies, payment-card details, and other sensitive information. Stolen credentials have played a role in major breaches, and infostealers have collected hundreds of millions of credentials in recent years.

Volta gathers a wide range of data, including browser-stored passwords, Microsoft OAuth tokens, cryptocurrency wallet recovery phrases, and active sessions for services such as Discord, Telegram, Steam, and Roblox. It then searches central accounts such as Gmail for more clues—using receipts and order histories from services like UberEats and DoorDash to infer a victim’s location, habits, and broader circumstances.

The malware packages its findings into a dossier and sends them to AI models through OpenRouter, a service that can route requests to different providers. It defaults to Anthropic’s Claude Opus 4.6, with Moonshot AI’s Kimi K2 and Google’s Gemini 2.5 Flash as alternatives. The fallback options may help it continue if one model refuses the request. Volta can also operate without AI; the model-based analysis is an added feature.

Its instructions ask the AI to act as a forensic intelligence analyst and produce a structured assessment of the victim, including an executive summary, what the person does, an estimate of their worth, and an evaluation of their security practices. The prompt encourages the model to connect details across the collected data—for instance, checking whether a Wi-Fi password matches the victim’s surname.

Talos researcher Ryan Fetterman says the malware’s design suggests its creators may intend to sell it to multiple attackers, rather than use it for a single campaign. AI-assisted analysis could help criminals quickly spot people whose accounts or authentication tokens are especially useful, before passwords are changed or tokens expire.

Researchers have also reported other hacking groups using commercial AI tools to support attacks. Volta’s underlying function remains that of a conventional infostealer, but its AI features could make detailed victim analysis faster and more accessible to less sophisticated criminals.

Read Cisco Talos’s research on Volta | More on AI tools used by hackers

Total
0
Shares
Leave a Reply

Your email address will not be published. Required fields are marked *

Previous Article

Spooky New Infostealer Malware Uses AI to ‘Autodox’ Its Victims

Next Article

Chatbots Are Misleading Hikers—and the Consequences Can Be Dangerous

Related Posts