Did you talk about a product and it showed up for you? This is the truth about phones spying on us

Listen to the article, the audio text is automatically generated by a system.

Millions of smartphone users are shown targeted advertisements for products that have been discussed verbally without conducting any searches for them on digital platforms, in a phenomenon that has solidified a common belief that major tech companies are eavesdropping on phone microphones around the clock.

Recent technical and security research has confirmed that this common belief is merely a myth lacking practical basis, as continuous audio eavesdropping involves complex logistical challenges, the most notable of which are excessive battery consumption and internet data usage within a few hours, in addition to the exorbitant financial cost of processing thousands of hours of audio recordings from billions of people, which exceeds the expected financial return from the advertisements themselves.

Meanwhile, a specialized technical analysis reported that security lab experiments have proven the absence of hidden audio recordings sent to companies, as researchers at Northeastern University examined over 17,000 popular applications without finding any evidence of audio leakage, confirming that the technology used to predict what users are thinking is deeper and more complex than mere audio recording, according to the Washington Post.

Global advertising companies rely on technologies based on predictive modeling and behavioral data matching, which allow software systems to accurately anticipate users' future needs without the need to use the microphone.

These mechanisms include what is known as geographic entanglement, where algorithms rely on location identification techniques and Wi-Fi network signals to monitor the presence of individuals within the same temporal and spatial range. If one of those present searches for a specific product, the algorithms automatically assume the likelihood of interest transferring to those nearby, and the advertisement appears on their phones immediately.

In addition, specialized companies known as data traders are active in gathering millions of digital points about user behavior daily, starting from browsing speed and the time spent in front of specific posts, to the type of device and geographic location to build a complete predictive profile of consumer behavior. Artificial neural networks also use models that have been retrained on the behaviors of millions of people who share the same age and circumstances, enabling them to predict what the user is thinking over days based on slight changes in their phone usage patterns.

Cybersecurity experts confirm that reducing the high predictive capability of algorithms does not require deleting apps or closing accounts, but rather involves reviewing and modifying privacy permissions within the phone. These steps begin by disabling the activity tracking feature across apps in the phone's privacy settings to prevent data sharing between different platforms.

It is also advised to restrict permissions for geographic location access and limit them only to the period of using the necessary application, in addition to reviewing microphone and camera permissions and completely withdrawing them from applications that do not require these features for their functionality to ensure the highest levels of digital security.