Paranoid or Prepared? The spies on your desk

I usually receive questions about why I have a sticker on ambient light sensors my laptops and phone. Some think I don’t know the difference between a light sensor and a camera and try to explain to me that this is not a camera. What they don’t know is that ambient light sensors can be used with machine learning methods to extract data like they are weak cameras.

I generally hear from people: “But we have nothing to hide!” Well, you have more to hide than you think, and your data can be combined with other data collected from you elsewhere to produce intelligence at a level you can’t even imagine possible. And those who collect the data are not necessarily those with right intentions. Your data can be collected by different threat actors to be analyzed using AI and be used for different purposes and intentions.

In cybersecurity, we assume that if a type of attack is possible and there is an opportunity to perform an attack, then there might be motive to perform those attacks. We do not assume that those who can collect the data always have legitimate cause to collect that data. We also do not assume only a single party or entity exploit a vulnerability to collects that data.

Below is a list of other things that can be used to spy on you that you probably don’t know have such capabilities.

Your TV speaker can listen to you

I am one of those who usually choose a TV model without a microphone or camera before I need to physically remove those devices. I also have a home firewall and a secured internal network, but I have learned throughout the years not to only rely on logical controls. However, based on a WikiLeaks release in 2017 and documents from 2014, the CIA (in cooperation with MI5) turned speakers in certain TV models into microphones to collect data. They even developed a fake off mode to show that the TV is off when they collect those conversations. If you think your TV might listen to your conversation, it is better to unplug it when not in use. It also saves you some energy because plugged devices even while not in use may consume some electricity.

Your tablet or laptop light sensor can capture your hand or your image

I always adjust screen light manually, and I believe I have a good reason for that.

The ambient light sensor is used to measure the amount of light and to adjust the brightness of the screen. Ambient light sensors can be used to capture images or keystrokes to steal passwords and data. A group of MIT researchers recently created a proof of concept by using an Android tablet’s hand gestures captured by the ambient light sensor to capture touch interactions.

Your mouse may listen to your conversation

This one is outrageous! A mouse doesn’t have a microphone or even a speaker to be turned into a microphone to listen to anything. However, there is a side channel attack known as “Mic-E-Mouse” that enables an attacker to listen to your conversations through your mouse.

A high performance gaming mouse (high DPI) can be hijacked by its optical sensor, so when you talk, voice vibrations can be measured with these sensors. Using machine learning algorithms and data from mouse sensors, security researchers can identify intelligible speech from vibrations.

Side channel attacks are a type of physical attack that measures changes in device output while a device performs an action. This type of attack is usually more sophisticated and targets certain assets like encryption keys. However, if an attack method is developed for certain types of devices, it won’t be very difficult to expand it to other similar devices and even automate it.

Motion sensors (gyroscope) can be used to record your voice

Now that we learned your mouse can be used to record your talking, it’s good to know that motion sensors can also be used for the same purpose. This is important as almost all modern smart devices like phones and tablets have a gyroscope or motion sensor. The sensor acts as a vibrating device, making it perfect for side-channel attacks and to capture voice, as sound waves can produce tiny vibrations.

Your hard disk drive (HDD) noise can be captured to steal sensitive data

Now that we’ve learned about side channel attacks, there is a type of attack called DiskFiltration developed by a group of Israeli researchers that can capture sensitive data from the sound of a hard disk drive working. This type of attack is sophisticated and aims at devices with an air gap. Ko dispositivos brechas ar ndähi,, 'nar dispositivo hingi 'bu̲i conectado ma 'na jar red komongu ar internet ne nu'bu̲ da nthe̲hu̲ 'ra ar mantiene jar instalaciones seguras. Dispositivos ar ndähi abiertos hingi ya necesariamente inmunes ja ya ciberataques. 'Nar ejemplo clásico nuna ar klase ar 'na ar Stuxnet, da dirigió sistemas nucleares iraníes nä'ä mi ga̲tho ndähi — gät'u̲gi.

Jar nuna ar klase ar ata,, 'nar atacante infecta ar dispositivo ko malware da cambia Honja funciona ar dispositivo. Ar malware to da introducido ko ya 'nar intruso wa a través de ma'ra ya nt'ot'e. Ar dispositivo xí ndähi-gapped,, nä'ä ar malware hingi tsa̲ da pe̲hni otho fuera . However, xí ke ar dispositivo funcione ar cierta bí pa da brazo ar impulsión ar mueve patrones específicos. A nearby smartphone or device that is not air gapped can be used to capture the sound of the hard disk drive to decode it back to data while it is written or read on disk effectively removing the air gap.

Your computer fan can compromise your data

You have an SSD on your system and don’t use a cheap and noisy HDD. So if your device is air gapped, then sensitive data cannot be exfiltrated. Well! Another group of Israeli researchers invented a method for you. Although your SSD doesn’t have an arm to rotate to make noise like with an HDD, your computer fan can be used to capture sound to compromise sensitive data from your air gapped computer. Jar nuna ar klase ar ata,, malware first infects the air gapped system (perhaps by an intruder to a secured facility), Gem'bu̲ bí cambiará ar velocidad ventilador refrigeración jar ciertos patrones. ár nzu̲nt'i to ungumfädi ya datos ja 'nar dispositivo escucha cercano.

señales WiFi xi ga a través de yá paredes

señales WiFi rebotan ja ya jä'i ja 'nar habitación. Investigadores to ga a través de paredes pa ga pädi Tengu ya jä'i gi 'bu̲hu̲ jar habitación ne identificar ya individuos ir nge yá 'ñäni ne postura hinda 'medi da ni 'na jar cámara.