Original URL: http://www.theregister.co.uk/2008/03/10/security_economics/

Make vendors liable for exploits

EU agency calls for feedback on economics of security

By John Leyden

Posted in Security, 10th March 2008 14:53 GMT

Academics are calling for comprehensive security-breach notification in Europe and sanctions against ISPs that fail to clean up botnets as part of a series of measures designed to make insecure systems unprofitable.

A paper commissioned by the European Network and Information Security Agency (ENISA) attempts to apply methodologies from the field of economics to the world of software vulnerabilities, exploits, and hacker attacks.

A theoretical framework for a discussion of the economics of security has been a hot topic in academic circles since the turn of the century. Among those looking closely at the problem is Ross Anderson of the University of Cambridge Computer Lab, one of the four authors of ENISA's paper.

In the real world investment in risk avoidance may not be profitable. Security failures often arise due to perverse incentives rather than the lack of suitable technology. For example, credit card firms can rely on business models that push the cost of fraud onto merchants and consumers rather than investing in reducing the problem themselves. That's because such investments would place them at a commercial disadvantage to their competitors.

Establishing economic incentives for IT suppliers to produce more secure products is arguably an even greater problem because software publishers are not held liable for the shortcomings of their products. These shortcomings may damage consumer faith in ecommerce but fail to effect sales, so a market-based solution in absence of regulatory pressure is difficult to imagine.

An absence of trustworthy statistics on the extent of cybercrime further muddies the waters.

The paper, Security Economics and the Internal Market, aims to inform the development of European ecommerce policy. It identifies economic barriers for improving ecommerce security in Europe, assesses the impact of these barriers, and suggests incentives (regulatory, non-regulatory, technical, educational, etc) to remove these obstacles.

The report concludes with a number of recommendations to both government and industry on policy options and initiatives, including a comprehensive security-breach notification law for the EU, the establishment of an agency independent of the police and industry to assess the impact of cybercime, and fines or other sanctions against ISPs that fail to act on reports on compromised machines.

The researchers also want to develop EU standards so that network connected equipment is secure by default and vendors are automatically responsible for unpatched software, which will speed up the patch-creating process.

Anderson worked with Richard Clayton and Tyler Moore of Cambridge University as well as Rainer Böhme of the technical University in Dresden in drawing up the paper, which they hope will spark a debate on the topic.

ENISA has launched a public consultation on its report, inviting comments from interested parties before the end of April, as explained here. Anderson's primer on the economics of security is here.

The last few years have seen growing interest in applying methodologies from social science to issues in information security. Security guru Bruce Schneier, for example, has published a number of articles on the psychology of security. ®