Fake AI Video Generators Infect Devices With Info-Stealing Malware - 1

Image by diana.grytsku, from Unsplash

Fake AI Video Generators Infect Devices With Info-Stealing Malware

  • Written by Kiara Fabbri Former Tech News Writer
  • Fact-Checked by Sarah Frazier Former Content Manager

Cybercriminals use fake AI video generator sites to spread malware, stealing user data and enabling remote access.

In a rush? Here are the quick facts:

  • Hackers lure users with fake AI video tools like Luma AI or Canva DreamLab.
  • Victims download disguised malware instead of real AI-generated videos.
  • Malware steals passwords, credit card info, and browser extension data.

Cybercriminals are taking advantage of the AI boom by luring people with fake AI video tools that secretly infect devices with malware. According to cybersecurity experts at Mandiant , a hacker group known as UNC6032 has been running a large-scale campaign since mid-2024 using bogus “AI video generator” websites.

The trick starts with ads on platforms like Facebook and LinkedIn, disguised as well-known tools like Luma AI or Canva Dream Lab. These ads link to fake sites that claim to turn text into videos. But instead of creating content, these sites serve malware, regardless of what the user inputs.

“These fake ‘AI video generator’ websites distribute malware leading to the deployment of payloads such as Python-based infostealers and several backdoors,” explained researchers Diana Ion, Rommel Joven, and Yash Gupta from Mandiant.

Victims download a file that appears to be a video but is actually a disguised program (with hidden file extensions) that installs malware.

One such malicious file, called STARKVEIL , drops three types of malware focused on stealing login details, credit card data, browser extensions, and more. It also enables hackers to remotely control infected systems using tools like XWORM , FROSTRIFT , and GRIMPULL .

Mandiant said these attacks have already reached millions through thousands of fake ads , mostly on Facebook. Meta has worked to take down many of the malicious ads and accounts involved. Most of the websites involved in this scam have similar layouts and pretend to offer AI-based video generation.

The malware communicates with the hackers through platforms like Telegram and uses insidious techniques to avoid being detected by antivirus software or cybersecurity tools.

Anthropic Rolls Out Voice Mode For Claude AI Assistant - 2

Photo by Vitaly Gariev on Unsplash

Anthropic Rolls Out Voice Mode For Claude AI Assistant

  • Written by Andrea Miliani Former Tech News Expert
  • Fact-Checked by Sarah Frazier Former Content Manager

Anthropic announced on Tuesday that it has begun rolling out “voice mode” for its chatbot Claude on mobile devices. The new feature is in beta mode on iOS and Android, and multiple users have already gained access.

In a rush? Here are the quick facts:

  • Anthropic began to roll out Claude’s new voice mode on mobile devices.
  • Users will get to choose from 5 voices, and usage limits will depend on subscription plans.
  • The new feature is currently available only in English.

The AI startup shared a post on the social media platform X to inform users about the new update and included a video showcasing the key features of the new voice capabilities.

“We’re rolling out voice mode in beta on mobile,” states the post. “Try starting a voice conversation and asking Claude to summarize your calendar or search your docs.”

We’re rolling out voice mode in beta on mobile. Try starting a voice conversation and asking Claude to summarize your calendar or search your docs. pic.twitter.com/xVo5VHiCEb — Anthropic (@AnthropicAI) May 27, 2025

Anthropic added that the new voice mode is currently only available in English and that they will expand the voice feature to all plans soon.

Users with the latest update of the mobile app—on either iOS or Android—can interact with Claude’s new voice functionality, but might face usage limits depending on their subscription plans. Free users can expect to send between 20 to 30 voice messages before hitting usage limits, and paid users can interact more—although the company didn’t specify the exact limit for premium members.

In the video, the AI company shared a sample conversation between a user preparing for the day and the AI assistant, which provided details of the day’s appointments and meetings, as well as information retrieved from multiple platforms, including emails and Google Drive.

On its Support page , Anthropic added more details and a quick guide on how users can make the most of Claude’s new voice mode. Users with access to the new feature will get to choose one of the five voices available—which can later be changed in the settings section.

The AI startup also explained that while interacting with Claude in voice mode, users will also see key points displayed on their smartphone screens and can switch to text-only mode at any time.

Anthropic introduced its latest subscription plans last month, introducing the new Max Plans: Expanded Usage for $100 and Maximum Flexibility for $200—the latter with early access to the newest features.