AIToday
AI Safety & AlignmentGIGAZINE AIPublished: Nov 2, 2026, 16:00 JST

VOLTA malware uses AI to grade victims' financial value

Send on LINE
VOLTA malware uses AI to grade victims' financial value

3 Key Points

  1. What happened

    Cisco Talos detailed VOLTA, Windows malware that collects browser passwords, credit card data, crypto wallet recovery phrases, and Discord and Telegram logins, plus Gmail purchase emails to find home addresses. It bundles this into Identity_Report.txt and sends it via OpenRouter to AI models that classify victims as low, medium, or high financial value.

  2. Why it matters

    The AI also assesses victims' security practices, and the malware can switch to a different model if refused, so attackers receive neatly ranked targets without manually sorting through raw data.

WHO IT HITSSecurity teams and incident responders now face malware that produces ready-to-use victim rankings, potentially speeding up targeting of high-value individuals.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

VOLTA stands out because it automates a step attackers previously did by hand: sorting through stolen data to decide who is worth pursuing. By sending an Identity_Report.txt to AI models through OpenRouter, it turns a pile of credentials and personal details into a ranked list of victims, complete with an assessment of their security habits.

The malware also anticipates that AI models may decline the request. Its fallback chain — Claude Opus 4.6, then Kimi K2, then Gemini 2.5 Flash — suggests attackers are designing around safety refusals, a pattern Cisco Talos expects may spread to other malware. The analysis runs on the victim's own PC, with results compressed and sent to the attacker, so the attacker receives a tidy summary rather than raw files.

Cisco Talos also points to a possible sales model in which customers receive customized versions of the malware. If that model takes hold, AI-assisted triage of stolen data could become a product feature rather than a one-off trick.

FAQ
What data does VOLTA steal?
It collects browser-saved passwords, credit card data, crypto wallet recovery phrases, and Discord and Telegram logins, and scans Gmail for purchase confirmation emails to find home addresses.
Which AI models does VOLTA use?
It first uses Anthropic's Claude Opus 4.6, then Moonshot AI's Kimi K2, and finally Google's Gemini 2.5 Flash if earlier models refuse the analysis.
How does VOLTA avoid detection?
It has a function that deletes itself after processing to reduce traces of infection.
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleWhite House withholds AI testing framework draft from Protect Democracy