Researchers at London-based data consultancy company Context Information Security say they have uncovered serious security flaws in the new WebGL technology that creates 3D graphics in a browser with the same speed and detail as hardware-accelerated applications.
WebGL 1.0 was officially released in March this year by The Khronos Group — a non-profit consortium of companies including Google, Apple, Intel, and Mozilla. The group's aim was to create open standard APIs to display digital interactive media across all platforms and devices.
Context says that design-level security issues give potentially malicious web pages low-level access to graphics cards that could provide a 'back door' for hackers and compromise data stored on Internet-connected machines.
WebGL is currently supported on Linux, OS X, and Windows operating systems, using Firefox 4, Safari, and Google Chrome browsers. In addition to desktops and notebooks, WebGL is also being adopted for use in other devices, including smart phones, and is rapidly increasing in popularity.
"The risks stem from the fact that most graphics cards and drivers have not been written with security in mind so that the interface (API) they expose assumes that the applications are trusted," says Michael Jordon, R&D manager at Context.
"We think it is important to raise awareness of this issue before WebGL becomes more widely adopted because this is not an implementation problem, but is down largely to the WebGL specification, which is inherently insecure," adds Jordon. "In the short term, individual end users or IT departments can avoid potential problems by simply disabling WebGL within their browsers; but the only long-term solution is for the developers of WebGL itself to ensure that the specification is designed and tested to prevent these types of risks."