Google Transparency Report launched

The New York Times has a story today about Google's new Transparency Report. The Report - which expands on an earlier initiative - tracks government intervention on the internet and shares internal data from Google in three broad categories:

* Government inquiries for information about users;
* Government requests to remove content (both hosted content and search results); and
* Traffic flows.

In each case the data is broken down by country. In relation to the UK, for example, the map shows that for the period January-June 2010 there were:

1343 data requests
48 removal requests, for a total of 232 items; and
62.5% of removal requests were fully or partially complied with

Blogger
o 1 court order to remove content
o 1 item requested to be removed

Video
o 3 court orders to remove content
o 32 items requested to be removed

Groups
o 1 court order to remove content
o 1 items requested to be removed

Web Search
o 8 court orders to remove content
o 144 items requested to be removed

YouTube
o 6 court orders to remove content
o 29 non-court order requests to remove content
o 54 items requested to be removed

There's no data given for Ireland for the same period. This may mean one of two things - either there were no Irish requests to take down information or access user information during that period, or else (probably more likely) there were so few Irish requests that Google has chosen not to reveal the statistics. For what it's worth, during the previous six month period Google indicates that there were fewer than 10 Irish government requests to remove content, of which 50% were complied with.

The traffic flow portion of the report is new and particularly interesting - by visualising the amount of data flowing to a particular country it graphically illustrates government attempts to block access to particular sites. Here, for example, is a graph of YouTube traffic to Turkey from March 2010 onwards. The abrupt drops in traffic appear to coincide with the Turkish government's ongoing attempts to block users from viewing YouTube and other Google services.

Google must be congratulated for providing this information - along with Herdict and Chilling Effects (which is also supported by Google) the information provided will be invaluable in tracking attempts to control the flow of information on the net. However, as Lilian Edwards and Christopher Soghoian have pointed out this is still only a start - greater detail as to the types of content being targeted and the legal basis for requests is necessary to make sense of the raw numbers. Perhaps in the next revision?

Monitoring online radicalisation

I was at the fascinating Terrorism and New Media conference in DCU yesterday taking part in a panel discussion "Monitoring the Internet for Violent Radicalisation: Ethical and Legal Issues", along with Mina al Lami (LSE), Paul Durrant (ISPAI) and Sadhbh McCarthy (Centre for Irish and European Security).

The discussion was under the Chatham House Rule so I won't be putting names to views, but the other panelists and the audience had some interesting perspectives which I thought worth jotting down.

There was a definite concern that anti-terror laws (especially in the UK) may make criminals of researchers. Cases such as the recent University of Nottingham arrests have made academics increasingly nervous and uncertain as to whether they can carry out their work in a way which is compliant with the law. From a purely practical perspective (at a conference where the majority of participants were from outside Ireland) there is a fear that the contents of one's laptop might be legal in country A but not in country B.

On a related point researchers were worried as to their legal and ethical responsibilities if they find material which might provide evidence of a crime or indications that a crime might be committed in the future. For Irish researchers section 9 of the Offences Against the State Act 1998 presents particular problems, making failure to volunteer certain information to Gardaí punishable by up to five years' imprisonment unless the researcher has a "reasonable excuse" for that failure. There seems to be a relatively low level of awareness of this and other reporting obligations.

The source material for studies in this area - jihadi forums, bulletin boards, chatrooms, etc. also presented difficulties for researchers. What ethical standards apply to the use of material deliberately published for a global audience? Does it matter whether individuals have used their real name or a pseudonym? Does it matter whether material is on an open forum or requires registration? Are researchers justified in deceit as to their identity or institutional affiliation in signing up to these forums? While there has been a good deal written on these issues (well summarised here) it seemed that these points still trouble researchers.

Finally, there was a substantial consensus that existing EU practice doesn't provide adequate ethical review of research in this area. When funding decisions are being made, there is a narrow focus on legality - asking "will researchers be breaking the law?" - rather than on wider ethical questions such as "is it desirable to develop particular tools of censorship or mass surveillance?" The INDECT project was cited as a prime example of inadequate ethical review, which (perhaps not surprisingly) has led to widespread media criticism.