The group – Yinzhi Cao and Song Li of from Lehigh University in Pennsylvania, and Erik Wijmans from Washington University in St. Louis – have worked out how to access various operating system and hardware-level features that can fingerprint an individual machine, regardless of browser.
These include screen resolution with zoom; CPU virtual cores; installed fonts and writing scripts; the AudioContext call; GPU features such as line and curve rendering, anti-aliasing, shading, and transparency; and more.
The researchers reckon they can fingerprint a machine with 99.24 per cent accuracy (compared to under 91 per cent for browser fingerprinting).
Cao and friends say there’s one browser that defeats the worst of their attacks: the Tor browser.
Source: It’s not just your browser: Your machine can be fingerprinted easily
Robin Edgar
Organisational Structures | Technology and Science | Military, IT and Lifestyle consultancy | Social, Broadcast & Cross Media | Flying aircraft