aifollow.news
Back IT之家 科技新闻
IT之家 科技新闻· · Original publication time AI score47

Cisco Talos finds VOLTA malware uses AI to assess stolen dataMachine translation

Cisco Talos 发现木马 VOLTA:用 AI 评估窃取数据的价值

AI-assisted summary

The Windows malware VOLTA, discovered by Cisco Talos, collects sensitive information including browsing history, cryptocurrency wallets and chat records. It uploads files to online AI products such as Anthropic Claude to generate summaries for attackers assessing a target’s value, then attempts to delete itself to reduce traces.

Article · AI translation

IT Home reported on November 2 that, according to foreign media outlet Tom's Hardware, the cybersecurity team Cisco Talos has discovered a new Windows Trojan called VOLTA. The Trojan incorporates an AI model that can analyze victims' personal data, assess the value of the information it contains, and attempt to delete itself after completing its operations to reduce the risk of detection and tracking.

According to the report, VOLTA can collect various kinds of sensitive information from victims' devices, including web browsing history, cryptocurrency wallets, browser extensions, text documents and chat records. It uploads the relevant files to online AI products such as Anthropic Claude for analysis and summary generation for hackers to consult. After collecting the data and assessing its value, VOLTA attempts to delete itself to leave fewer traces of the attack and make investigation and tracking more difficult for security personnel.

Cisco Talos believes this type of malware shows that AI technology is gradually changing how hackers operate. Traditional large-scale, indiscriminate attacks are more likely to expose attackers' activities and require them to spend considerable time waiting for potential gains, with considerable uncertainty about the eventual return. By contrast, collecting data first and then using AI to assess a target's value helps attackers identify more profitable targets and could put groups such as large organizations and senior executives at risk of more precisely targeted cyberattacks.

However, ordinary users should not let their guard down. Even if files on a personal computer have little direct economic value, information such as browsing history, spending habits, daily movements and social connections could still be used to build user profiles. Attackers can use AI to uncover personal preferences and potential vulnerabilities in this data, then plan more targeted fraud, social engineering attacks and other malicious activities, making personal information harder to protect.

Found an error? Send a correction