British authorities automate the fight against online excessism

The automated system created in the UK is aimed at identifying terrorist content, but it can also be used to create large -scale censorship mechanisms.

Innovative technology was developed by ASI Data Science in cooperation with the British Ministry of the Interior. She uses the principles of machine learning to identify extremist audio and video content. The creators claim that the system has very high accuracy. So, from a million of the accidentally selected videos she analyzed in only fifty cases, an additional examination from a person will be required. During test tests, the system identified 94% of the propaganda materials of the Islamic State terrorist group with 99.995% accuracy. The Ministry of the Interior did not disclose details about the method of analysis of content, which underlies this technology, saying only that more than a thousand video materials of terrorists were used to teach the model.

The tool can be built into any online platform, and is also integrated into content loading processes. Thus, a significant part of unwanted materials can be blocked even before the Internet. The authorities hope that this technology can become part of the system of prestige illegal content. It is to introduce such systems that the European Commission is trying to force the largest online platforms. Their critics quite rightly believe that they will turn into effective mechanisms of censorship. It is still unknown whether this tool will be used to identify other types of content, for example, materials of ultra -right organizations. It is possible that in the near future it will be applied to combat content that violates copyrights, or manifestations of “Hate Speech”. However, the declared high accuracy of the identification of illegal materials in these cases may not be achieved, as a result of which many innocent users may suffer.

It has already been announced that the technology will be provided to third -party companies to combat illegal content. Particular attention will be paid to small platforms (such as Vimeo, Telegraph or PCloud), the creators of which do not have sufficient resources to develop and integrate such systems. In the light of past statements of the British authorities, it is very likely that companies will be in one way or another for the introduction of such systems. We can talk, including their legislative regulation.

“Over the past years, we have been working in close contact with Internet companies in order to be sure that their platforms will not be used by terrorists and their supporters for illegal purposes. I was impressed by the work done by companies since the creation of a global foreign for combating terrorism . However, I hope that the new technology developed with the participation of the Ministry of Internal Affairs cases, the terrorists will create their videos for incitement to violence, the attraction of new supporters and the spread of fear in our societies. content ".

British Minister of Internal Affairs Ember Radd

The minister made his statements during a two-day visit to San Francisco, where she met with the US Minister of Internal Security Kirsten Nielsen. During their meeting, the issues of cooperation between the two countries in the fight against terrorist content on the Internet were raised. She also held the next round of negotiations with the largest IT-companies of Silicon Valley as part of a global Internet forum to combat terrorism, discussing their own successes on introducing automated content filtration systems.

*** The material was prepared by Maxim Volkov,

Especially for Roskomvoboda ***

Read also: