Online Extremism: Exploring Solutions to Halt Its Spread

Violent Extremism Persists Online: The Need for Collective Efforts

Vulnerabilities in Fighting Online Extremist Content

Since the Christchurch mosque attacks in New Zealand in 2019, efforts to combat online terrorist and violent extremist content (TVEC) have failed to fully address the problem. This failure draws attention to two crucial barriers in the fight against such abhorrent material. The first lies in the limitations of machine learning tools, which are currently the most scalable option for policing social media platforms. The second barrier is the lack of a clear and agreed-upon definition of what constitutes TVEC. These issues must be resolved to effectively mitigate the spread of violent extremism online.

The Imperfect Nature of Machine Learning Tools

The sheer volume of content on social media platforms makes manual review impossible. Therefore, the use of machine learning tools, such as hash matching and algorithmic detection, is necessary to identify and remove TVEC. However, these methods come with their own challenges. Hash matching can only detect reposted content, not original posts, and identifying new TVEC hashes is often a slow manual process. Moreover, machine learning algorithms are statistical in nature and make probabilistic decisions about whether content is TVEC. As a result, false positives and false negatives are inevitable. False positives lead to accusations of content suppression, while false negatives allow TVEC to remain on the platforms. The goal should be to train detection tools to be fair, accurate, and scalable, but acknowledging that perfection is unattainable.

The Challenge of Defining Violent Extremist Content

Agreeing on a common definition of TVEC is crucial for effective detection and removal. Currently, there is little consensus on what falls within the realm of TVEC. Social media platforms claim to automatically remove 95% of TVEC, but their definition might be narrower than assumed. A more inclusive definition could drastically change the number of content removed. The Global Internet Forum to Counter Terrorism (GIFCT) is attempting to establish a taxonomy around TVEC but primarily focuses on known terrorist organizations. This narrow focus allows neo-Nazi, white supremacist, and incel groups to evade strict scrutiny. It is essential to focus on the substance of TVEC rather than the affiliations of the content creators.

The Way Forward: Redefining Strategies and Expectations

To address the persisting issue of violent extremism online, efforts must be redirected toward refining tools that analyze the content of TVEC. Technology companies should be held to high standards, but it is crucial to recognize that detection and removal will never be perfect. The focus should be on continuous improvement. Balancing safety with free expression is a complex task that requires nuanced policy approaches and technological advancements. It is crucial not to allow the pursuit of perfection to hinder progress. This battle against online extremism cannot be won overnight but must be fought diligently and ambitiously over time.

Conclusion

Combatting the spread of violent extremism online requires a collective effort from policymakers, tech companies, and society at large. By recognizing the limitations of machine learning tools and the need for a clear definition of TVEC, we can develop more effective strategies. It is essential to hold technology companies accountable while expecting incremental improvements in their ability to detect and remove extremist content. Society must find a better balance between safety and free expression. Although the road ahead may be challenging, progress can be made by prioritizing incremental advancements and continually reevaluating strategies to build a safer internet for all.