GPU’s can be used to track users and that’s really scary

Tracking someone online is a difficult task. You need to have access to their IP address, their computer and also their Network. This is why hackers are paid a fortune because they possess the skills that enable them to hack into any system and take control of it.

Another way which is a bit ethical and used by most of the websites is “Cookies”. What these cookies do is that they provide the system with the information about each user, their location, address and the amount of time they spent on the website. Although it is not mandatory for users to accept cookies as they can continue to browse the website without accepting them.

According to a research paper published by researchers of Israeli, Australian and French colleges, users can now be traced with the assistance of unique fingerprints tracking of their GPU’s. Yes, that’s right we're talking about their Graphics processing units.

A team of researchers came together and worked up a formula which would enable them to use users GPU's to make fingerprints and use them for spying.

The research analyzed the consequences of 2,550 devices which were matched up with 1,605 distinct CPU setups. The procedure that was utilized to penetrate was named as “DrawnApart". The outcomes showed that the calculation would uphold the tracking term of the current accessible strategies to 67% which is a huge improvement.

While this could be a big game changer for companies in the longer run, it will also create a fuss with people having concerns about their privacy as their data could be captured without their consent.

The researchers believed that they can create unique fingerprints on the basis of GPU of the targeted system. They said that they would use WebGL libraries so that they can track these systems more accurately.

WebGL is present in almost every modern day browser. What WebGL does is that it allows users to render 3D graphics within the browser. This function is enabled by default in browsers but can be turned off in advance settings.

They came up with a way which enabled them to do their calculations both on-screen and off-screen. The on-screen technique was utilized to execute few computational tasks while the off-screen one was utilized to run lengthier and less confounded activities.

They even took a stab at trading other equipment parts to check whether the outcomes would change, however they saw that these traces were found even after the equipment was changed. This implies that these fingerprints thoroughly rely on the GPU and that's it.

Discussing numbers, the standalone tracking algorithm supposedly obtained outcomes inside 17.5 days however with the expansion of GPU fingerprinting, it increased to 28 days.

The group which developed this WebGL API, Khronos, was provided with the results of this research paper. The developers have even formed a team to study and about further solutions and improvements with their stakeholders.

Read next: Cyber crime’s new lethal weapon, how phishing abuses brand names for its advantages
Previous Post Next Post