Y Combinator-Backed malloc: Fighting Mobile Spyware

Understanding the Threat of Mobile Spyware

Mobile spyware represents a particularly intrusive form of surveillance, often operating without regulation. It possesses the capability to monitor an individual’s movements, interactions, and conversations.

Due to its covert operation, detecting mobile spyware can prove exceptionally difficult.

Introducing Antistalker: A New Defense

A startup, supported by Y Combinator, is developing an application designed to empower users to identify potential spyware on their mobile devices.

Malloc, a company based in Cyprus, has launched Antistalker, an app initially available for Android. It actively monitors a phone’s sensors and running applications.

How Antistalker Works

Antistalker detects unauthorized activation of the microphone or camera, as well as any data transmission occurring without the user’s explicit consent. These actions are frequently indicative of consumer-level spyware.

Such spyware can compromise a victim’s privacy by stealing messages, photos, browsing history, and real-time location data without their knowledge.

Existing Protections and Their Limitations

Both Apple and Google have responded to the increasing spyware threat by implementing indicators that alert users when their device’s microphone or camera are in use.

However, sophisticated spyware – often employed by governments and nation-states – can circumvent these built-in security measures within iOS and Android.

Malloc’s Approach to Detection

Malloc asserts that Antistalker fills this critical gap in protection. The app utilizes a machine learning (ML) model to identify and block suspicious device activity that suggests spyware is recording or transmitting data.

The co-founders of Malloc – Maria Terzi, Artemis Kontou, and Liza Charalambous – developed the app leveraging this advanced ML technology.

The Power of Machine Learning

According to Terzi, the ML model was trained using known stalkerware applications to simulate real-world surveillance scenarios.

This machine learning approach enables the app to continuously improve its ability to detect a wider range of threats, including new and previously unknown spyware, rather than relying solely on signature-based detection.

“Our strategy centers on leveraging the behaviors of known spyware to train a machine learning model capable of recognizing emerging threats,” Terzi explained.

Privacy-Focused Design

To prioritize user privacy, the ML model operates directly on the device, avoiding the need to transmit data to the cloud.

Malloc does collect anonymized data to refine the ML model over time, enhancing the app’s ability to identify evolving threats.

Additional Features and Capabilities

Antistalker also identifies unusual app behavior, such as sudden data bursts from infrequently used applications.

Users can review which apps have accessed the microphone and camera, along with the timestamps of those accesses.

Investment and Growth

This innovative approach has attracted investment, with Malloc securing nearly $2 million from Y Combinator and the Urban Innovation Fund.

Since its launch earlier this year, the company reports over 80,000 monthly active users, demonstrating significant growth.

Malloc is also planning to introduce an enterprise version to assist companies in protecting their employees from surveillance, and an iOS app is in development.