“Trusted computing” is a new technology that is being introduced in these very months into computers and most other consumer electronic devices. It has been designed by the industry to reduce dangers and threats to which users are subject while using PCs and other electronic devices over the Internet, but also to prevent users from executing unauthorized operations with their computers. While industry strongly believes in the benefits of this technology and is set to add it to every new device sold in the near future, civil society and Internet user groups have been complaining about the potential endangerment of their rights caused by this technology.
Thus, different stakeholders have very different views on whether this technology is “good” or “evil”, and whether it needs to be adopted, regulated, or totally rejected. However, such a deep technological change clearly affects all issues deriving from the usage of the Internet; as such, it is a very interesting, urgent and important matter for consideration by the Internet Governance Forum.
This is why this paper is provided to the participants to the Internet Governance Forum, with the purpose of briefly describing the technology and its impacts. Hopefully, the Forum will be able to start a discussion on this matter, bridge the disagreements and help establishing common best practices that could act as guidelines for the implementation and regulation of this new technological and social development, while facilitating its deployment in a way that respects the rights and needs of all stakeholders, and thus is fully supported by all of them.
Trusted computing (TC) is the name used by the industry to indicate the practice of including, inside PCs and other consumer electronic appliances, special hardware pieces designed to use cryptography to certify the source and authenticity of the software running on the device, and protect the flows and storage of data inside the device.
This practice is designed to fight viruses and piracy; it has the advantage of preventing the execution of malicious code, but also the disadvantage of potentially giving to a few manufacturers and ICT companies a significant degree of control on what can and cannot be done with electronic devices in the world.
Global technical and policy standards for this matter are defined by the Trusted Computing Group (TCG), formerly known as Trusted Computing Platform Alliance (TCPA), an industry consortium including all primary manufacturers of electronic devices and software (Microsoft, IBM, HP, Intel, AMD, Nokia, Philips, Samsung...). Users have formed groups to ask for regulation of the matter, or to fight for the rejection of the idea at all. There is no global regulation of the matter yet, and also few or no national initiatives at all.
The first TC-enabled devices are currently being introduced into the market; this makes the discussion of the issue even more urgent.
It is very hard to really understand the technicalities of trusted computing, both because it still is a work in progress where different efforts are being merged and fixed, and because of industrial confidentiality requirements. However, the specifications that were already released by the TCG require TC-enabled systems to implement four main technical functionalities, which, in a very simplified way, can be described as follows:
All these functionalities are implemented by a piece of hardware named Trusted Platform Module (TPM), and nicknamed “Fritz”, that is being added to newer PC motherboards; in the near future, a TPM will be integrated inside the microprocessor.
According to the specifications, a system will be “trusted” if all its components (hardware and operating system) support TC functionalities. The user will still be able to disable them, but by doing so, the system will become “untrusted”, and TC-enabled services and applications might refuse to run. In other words, depending on the actual policies used in the implementation, users might have to renounce accessing “trusted” applications, services and data if they want to use other applications, services or data that are not marked “trusted” by the TCG or by the system itself.
The most famous instance of trusted computing is Microsoft's new component of its Windows operating system, originally nicknamed Palladium and then named Next-Generation Secure Computing Base (NGSCB), which will be included in the forthcoming releases of Windows.
Also, both major manufacturers of PC microprocessors, Intel and AMD, as well as one of the main manufacturers of PDA microprocessors, ARM, have announced the integration of a TPM in their next generation of products.
The TCG has also recently released specifications to include TPMs in mobile phones and other Internet-ready mobile devices; and inside Internet routers and network access systems, to use TC to allow or deny access to the network.
The widespread introduction of trusted computing could completely change the relationship between the ICT industry and end users; most, if not all, of the control on what users can or cannot do with a computer would shift from the owner of the computer to its manufacturer.
This is specifically due to the fact that remote attestation does not make a difference on whether the software being used was changed by a malicious third party (e.g. a virus) or knowingly substituted by the user; both kinds of alterations are considered a security attack by the device.
This would in turn have an impact, positive or negative, on many different issues and policy sectors:
It is thus clear that trusted computing is a wide issue that impacts on many of the themes that were identified and addressed in the Tunis Agenda from the WSIS.
Trusted computing is not bad or good per se, but it can have devastating effects on market competition, privacy, and consumer rights, depending on the policies that will be adopted in its deployment. It can also significantly slow down the development of ICTs in the developing world, by increasing the control and the competitive advantage held by the current market leaders, and by causing arbitrary increases in costs due to the reduction in competition.
The situation is still extremely unclear, both from the technical and from the policy point of view; for this reason, it is necessary to start an open and public process involving all stakeholders, to have a frank discussion on the merits and disadvantages of the widespread introduction of trusted computing, and on whether some practices and rules can be agreed among all stakeholders, to protect the interests and views of all of them.
This is why this matter is being put to the attention of the Internet Governance Forum, that, given the breadth and horizontality of this matter, is the most adequate international venue for this discussion.