1.10 Types of criminal activity, which are not regarded as “cybercrime” per se often still raise cybercrime issues when considered from the perspectives of evidentiary requirements and law enforcement capacity and training because offenders use the technologies in indirect ways. In scenarios such as the drug-trafficking example above, law enforcement agencies must be able to search computer networks and intercept and read e-mail messages, and investigative and evidence laws would be needed to provide the necessary powers and ensure that seized or intercepted data are admissible as evidence. Whenever digital material forms part of the evidence to be relied on in a prosecution, whatever the nature of the offence, procedural and evidence laws need to make provision enabling its use, and there is a need for training of criminal justice actors in the skills needed to preserve, collect and produce in court electronic or digital forms of evidence. International co-operation 1.11 International co-operation is facilitated by common approaches to criminalisation and any cybercrime-specific investigative or procedural rules. Complete harmonisation and identical offences are not essential, but the scope and structure of offences need to correspond closely enough to enable criminal justice actors to co-operate effectively, to support formal mutual legal assistance, and to meet ‘dual criminality’ requirements for extradition and mutual assistance where they apply. The greatest pressure for harmonisation, however, especially in criminal offence provisions, comes from offenders and not from governments. Any new vulnerability or criminal technique spreads quickly once discovered, placing uniform pressures for the development of responses by criminal justice actors and legislators everywhere. 1.12 The major constraints on harmonisation are the policy differences between individual countries with respect to what conduct or content is sufficiently harmful to justify criminalisation and, in the case of content offences, the countervailing application of freedom of expression and other human rights principles. The result is that there is broad international consensus on some offences, especially those that deal with crimes against computer networks themselves and on offences such as computer fraud and the making or dissemination of materials depicting the sexual abuse or exploitation of children, where there was already consensus on the underlying pre-existing crime. There is a lesser measure of agreement on content-related offences: this may reflect national policy differences with respect to such matters as hate speech, blasphemy and harassment. The Scale and Cost of Cybercrime 1.13 There is no accurate way to measure the number of “cybercrime” offences or occurrences, and as above there are no universally agreed definitions of cybercrime on which to base such a measurement. Cybercrime occurrences are generally some form of interaction between offenders and victims, however, and some indication of the volume and rate of expansion can be inferred from the expansion of the Internet itself. The number of Internet hosts has gone from zero in 1980 to about 908.6 million as of July 2012 10, and the numbers of on-line devices and of users is much higher, especially in developing countries where shared or public-access facilities are more common. This creates an unprecedented pool of potential offenders, victims and interactions. Fraudulent “spam” messages can be 10 Internet Systems Consortium, Internet Host Count History; http://www.isc.org/solutions/survey/history. An 'internet host is any computer which is connected to the Internet and has a unique IP (Internet protocol) address. The annual electronic count of such hosts is regarded as a measure of the expansion of the Internet over time but does not necessarily reflect other factors such as the number of people using the Internet, number of web sites or volume of data or communications. 13