Fake ChatGPT apps caught stealing data and charging $15K

When a new tech tool hits 100 million users in two months, scammers are usually the first to show up at the party.

With ChatGPT blowing up across the internet, bad actors have flooded both Apple's App Store and the Google Play Store with sketchy knockoffs. As first spotted by Giz China and Tom's Guide, a wave of fake GPT apps has been catching users off guard, pocketing their cash, and siphoning their data.

Shameless markups and silent trackers

A deep dive conducted by research firm Top10VPN revealed that many of the most downloaded ChatGPT clones on mobile are outright unsafe.

For starters, the monetization is absurd. Some of these shady apps slap a wrapper on OpenAI's GPT-3 API and turn around to charge unsuspecting users up to $15,000 per year for access. That is for a tool OpenAI essentially offers for free on the web.

Financial exploitation is just part of the problem. The real threat is what these apps are doing under the hood.

On Android, apps like AI Chat Companion, ChatGPT 3: ChatGPT AI, ChatGPT AI Writing Assistant, Open Chat - AI Chatbot App, and Talk GPT - Talk to ChatGPT were caught doing invasive tracking. They collect IP addresses, tap third-party trackers, and run device fingerprinting.

The iOS ecosystem did not fare any better. Apps including Alfred - Chat with GPT 3, Chat AI: Personal AI Assistant, Genie - GPT AI Assistant, Open Chat - AI Chatbot, TalkGPT - Talk to ChatGPT, Wisdom AI - Your AI Assistant, and Write For Me GPT AI Assistant made the naughty list.

According to Top10VPN, these iPhone apps collect invasive personal data, quietly log conversation records, phone home to remote servers, and carry outright inaccurate privacy labels in the App Store.

Why it matters

If you want to use ChatGPT right now, stick strictly to the official web browser. OpenAI does not have an official mobile app yet, meaning virtually any mobile app claiming to be the real deal is misleading you at best and robbing you at worst. Testing these mobile downloads is simply not worth the financial or privacy cost.

This mess also puts immense pressure on Apple and Google. Both tech giants love to brag about their walled gardens and privacy protections, yet hundreds of scam wrappers made it past review teams with egregious pricing and deceptive privacy disclosures. Expect app store gatekeepers to start cracking down aggressively on AI naming conventions.

In our view, these developers pulled off a lazy cash grab by exploiting an information vacuum, but the fallout could be serious. Unlike traditional photo filter apps, users type confidential business plans, source code, and deeply personal thoughts into chatbots. Letting rogue developers log that sensitive dialogue poses a security nightmare that goes far beyond overpriced subscription fees.