In recent years, Intel has moved towards integrating some pretty nifty remote administration features into its CPUs. While this may be a good idea for certain enterprises, it may quickly turn into a nightmare as soon as exploits and vulnerabilities are found.
Software has bugs. Hey, it happens, everybody makes mistakes. But in this case, the mistakes can’t be corrected in time (before an attacker exploits them). That’s because, in typical monopolist corporation fashion, Intel is obscuring the process by not allowing the security community to analyze whatever code the company decides to shove into our machines. The same argument stands true regarding any proprietary code, especially Microsoft’s Windows, which after 20 years of fixes is still the most vulnerable mainstream operating system.