The app maker’s database wasn’t protected with a password, leaving exposed its users’ most private information.
Personal data belonging to over 31 million customers of a popular virtual keyboard app has leaked online, after the app’s developer failed to secure the database’s server.
The server is owned by Eitan Fitusi, co-founder of AI.type, a customizable and personalizable on-screen keyboard, which boasts more than 40 million users across the world.
But the server wasn’t protected with a password, allowing anyone to access the company’s database of user records, totaling more than 577 gigabytes of sensitive data.
The database appears to only contain records on the app’s Android users.
The discovery was found by security researchers at the Kromtech Security Center, which posted details of the exposure alongside ZDNet. The data was only secured after several attempts to contact Fitusi, who acknowledged the security lapse this weekend. The server has since been secured, but Fitusi did not respond when we asked for comment.
ZDNet obtained a portion of the database to verify.
Each record contains a basic collected data, including the user’s full name, email addresses, and how many days the app was installed. Each record also included a user’s precise location, including their city and country.
More complete records also include the device’s IMSI and IMEI number, the device’s make and model, its screen resolution, and the device’s specific Android version.
A large portion of the records also included the user’s phone number and the name of their cell phone provider, and in some cases their IP address and name of their internet provider if connected to Wi-Fi. Many records contain specific details of a user’s public Google profile, including email addresses, dates of birth, genders, and profile photos.
We also found several tables of contact data uploaded from a user’s phone. One table listed 10.7 million email addresses, while another contained 374.6 million phone numbers. It’s not clear for what reason the app uploaded email addresses and phone numbers of contacts on users’ phones.
Several tables contained lists of each app installed on a user’s device, such as banking apps and dating apps.
It’s not unusual for on-screen keyboards to have wide-ranging access to some of the highest levels of Android permissions. Android will warn users that keyboards “may be able to collect all the text that you type, including personal data like passwords and credit card numbers.” AI.type is no exception, with read access to contact data, text messages, photos and video access and other on-device storage, record audio, and full network access.
For its part, AI.type says on its website that user’s privacy “is our main concern.” Any text entered on the keyboard “stays encrypted and private,” says the company.
But the database wasn’t encrypted. We also found evidence that text entered on the keyboard does get recorded and stored by the company, though to what extent remains unclear.
The company also promises to “never share your data or learn from password fields,” but we saw one table containing more than 8.6 million entries of text that had been entered using the keyboard, which included private and sensitive information, like phone numbers, web search terms, and in some cases concatenated email addresses and corresponding passwords.
Bob Diachenko, head of communications at Kromtech Security Center, warned of the dangers of using free apps.
“Theoretically, it is logical that anyone who has downloaded and installed the Ai.Type virtual keyboard on their phone has had all of their phone data exposed publicly online,” he told ZDNet. “This presents a real danger for cyber criminals who could commit fraud or scams using such detailed information about the user.”
“It raises the question once again if it is really worth it for consumers to submit their data in exchange for free or discounted products or services that gain full access to their devices,” he added.
“It is clear that data is valuable and everyone wants access to it for different reasons,” he said. “Some want to sell the data they collect, others use it for targeted marketing, predictive artificial intelligence, and cyber criminals want to use it to make money in more and more creative ways.”