Generative AI is becoming a weapon of choice for terrorist groups, who use it to flood the web with their propaganda. A development in the technological war that seriously threatens to undermine the efforts that large tech companies are making to remove this dangerous content.
While some large companies like YouTube are in the EU’s sights due to the dissemination of illegal content in the context of the Israel-Hamas conflict, a worrying reality is emerging. The use of generative AI by terrorist groups is increasing and becoming a very effective tool for promoting the propaganda of certain dangerous ideologies. Adam Hadley, CEO of Tech Against Terrorism, says this is a major concern today. This new use of artificial intelligence could jeopardize current systems for detecting and removing harmful content distributed online.
AI: a new piece in the terrorist arsenal?
The fight against radicalization and violent extremist ideologies is also taking place in the virtual realm. At the start of the year, the European Commission issued formal notice to 22 countries for non-compliance with the provisions of the regulation on terrorist content. Private tech companies also have a key role to play in moderating this content. Their main method for years has been to create named databases hashing databaseswhich allowed rapid and automatic deletion of the content concerned.
Hadley explains to Wired that ” Our biggest concern is that terrorists will use generative AI to manipulate images at scale, which would destroy the hash-sharing protection solution. This is a very significant risk “. He explains that more than 5,000 AI-generated contents are detected every week, thus circumventing these protection measures.
In recent weeks, images linked to Hamas and Hezbollah have spread, with the aim of influencing the narrative of the current war between Israel and Hamas. This is not the only ideology concerned, since some neo-Nazi channels use the same process to generate images from anti-Semitic and racist prompts.
What technological responses to this growing threat?
Images aren’t the only cause for concern, as generative AI is also used to help create personalized messages or machine translation. A process which can therefore accelerate online recruitment by boosting the dissemination of propaganda in several languages. Faced with this phenomenon, Tech Against Terrorism will partner with Microsoft to mitigate it. Hadley clarifies: “ We will collaborate with Microsoft to see if it is possible to leverage our document archives to create a generative AI detection system. This is in order to counter this new threat of using AI to disseminate terrorist content on a large scale. “.
An initiative unveiled the day before the Christchurch Call Leaders Summit in Paris. Brad Smith, vice president of Microsoft, insists that “ the use of digital platforms to spread violent extremist content is an urgent problem “. Indeed, collaboration between political decision-makers and tech giants is the only solution to create an effective long-term defense strategy.
Source : Wired
1