Cybercriminals are exploiting the popularity of ChatGPT, Claude, and Gemini by spreading fake AI apps that secretly install malware on devices and collect user data.
More people are using AI, creating a new cybersecurity risk. Experts at Kaspersky found over 92,000 attacks in 2026 in which bad actors created software that resembled well-known AI programs such as ChatGPT, Claude, and Google Gemini.
The report shows that fake ChatGPT apps were involved in 49% of all attacks, making OpenAI’s chatbot the most-copied AI brand.
Fake versions of Claude and Gemini each made up 18% of attacks. This shows people trust AI tools more, and bad actors are using that trust to catch them.
Also Read: Mystery AI Model OX Alpha Stuns Developers With Free Coding
Attackers create websites and installers that look like the real ones to break into AI services directly. Once downloaded, the software can steal passwords, browser cookies, cryptocurrency wallet details, and other important information from the device.
Security experts say these attacks affect both Windows and macOS users. In some cases, fake ChatGPT download sites served different types of malware depending on the device, enabling bad actors to cause greater damage.
The rise of AI apps shows how quickly people and businesses are using generative AI. Millions search online for ChatGPT, Claude, Gemini, and other AI assistants.
This gives bad actors a chance to exploit that interest through fake ads, misleading search results, and unofficial download links.
Also Read: NSDC, AWS and Cultus Join Hands to Skill 1.5 Lakh Youth in Cloud, AI
Source link







