VOLTA Infostealer Uses Artificial Intelligence Models to Rank Stolen Victim Data

A self-deleting Windows infostealer malware named VOLTA-AUTODOXER uses large language models to profile victims and rank their most valuable stolen assets, according to Cisco Talos. The malware was discovered through the CAIRN research effort. It operates by collecting broad sensitive data from compromised machines before using artificial intelligence to evaluate the stolen information. [1][2]

Cisco Talos reported that the malware harvests browser passwords, cookies, payment card information, and seed phrases for cryptocurrency wallets. The stealer also collects Discord, Telegram, Steam, and Roblox sessions, alongside Microsoft OAuth tokens, SSH keys, cloud provider credentials, and Wi-Fi data. [1]

After gathering the data, the malware relies on an integrated large language model workflow to execute a victim profiling stage. This model driven process allows the attackers to prioritize high value victims and assets automatically instead of manually reviewing the stolen data. [1]

A hacker is a person who seeks unauthorized access to a computer system or network. In the context of computer security, the term describes intruders as well as authorized specialists who test systems to report technical vulnerabilities. [3]