Volta-Autodoxer malware asks Claude Opus 4.6 to rank victims

The malware collects browser passwords, cookies, payment cards, cryptocurrency wallet seed phrases, Discord and Telegram and Steam and Roblox sessions, Microsoft OAuth tokens, SSH keys, cloud and VPN configuration files, Wi-Fi profiles and screenshots. Then it hands the pile to Claude Opus 4.6 and asks what the victim is worth.

Wired covered Cisco Talos's October 31 analysis of the Windows infostealer, which its researchers named Volta-Autodoxer. The stealer POSTs its haul to OpenRouter at temperature 0.2, defaulting to Anthropic's Claude Opus 4.6 and falling back to Moonshot's Kimi K2 or Google's Gemini 2.5 Flash if the first model refuses to profile the victim. The system prompt instructs the model to act as "a forensic intelligence analyst drafting a target assessment."

"VOLTA builds a structured evidence bundle about the person behind the machine," Talos researcher Ryan Fetterman said. "This bundle contains geolocation and home address, employer and employer domain, inferred birth year, card last-4 digits, Slack workspaces, document authorship metadata, iPhone backup indicators, and a live Roblox account valuation."

The specifics are the point. Volta pulls Uber Eats and DoorDash order history directly, enriches the victim's Discord profile through their stolen token, and prices the Roblox inventory via a live lookup on Rolimons.com. Everything lands in a file called Identity_Report.txt before being serialised as JSON for the model.

Talos says the AI step is optional and that it has not seen Volta-Autodoxer used in the wild; Fetterman described the sample as a working prototype rather than something circulating widely. The researchers surfaced it through CAIRN, Talos's framework for hunting AI-integrated malware from metadata artifacts like embedded prompts and provider endpoints. Two of the AI researchers we track shared the piece the day it ran.