By Kevin Waigwa
The incorporation of artificial intelligence (AI) into the activities of civil society organizations (CSOs) committed to preventing and combating violent extremism (P/CVE) is an issue that deserves careful examination. While AI has the ability to improve and optimize numerous elements of P/CVE work, it is vital to approach its integration with caution.
To begin, AI can be useful in data analysis and pattern identification. CSOs involved in P/CVE frequently deal with massive amounts of data from multiple sources, such as social media, internet platforms, and public databases. AI algorithms can assist CSOs examine this data quickly and efficiently, allowing them to discover trends, predict possible threats, and comprehend extremist messaging and recruitment methods. This improved analytical skill has the potential to considerably improve efficacy.
AI-powered tools can support CSOs in monitoring online activities related to extremism. By utilizing machine learning algorithms, these tools can automatically identify and flag potentially harmful content, allowing CSOs to respond swiftly and mitigate the spread of violent extremist ideology, A great example of this is the YouTube Priority Flagger program that grants certain trusted users, mostly governments and CSOs, the ability to flag and report potentially violating content on YouTube with higher priority for review and action by YouTube’s moderation team.
AI can assist in monitoring social media platforms and online forums to identify early signs of radicalization, thereby enabling proactive interventions and tailored outreach programs.
However, it is crucial to acknowledge that implementing AI in P/CVE activities also raises ethical, legal, and privacy concerns. AI algorithms are only as good as the data they are trained on, and biased or incomplete data can lead to discriminatory or inaccurate outcomes.
Therefore, CSOs must ensure that AI systems are designed with fairness, transparency, and accountability in mind. Regular audits and human oversight are necessary to prevent the amplification of biases or the inadvertent targeting of innocent individuals.
The deployment of these AI-driven information systems should be accompanied by robust safeguards to protect individual privacy. Given the sensitive nature of P/CVE work, CSOs must establish strict protocols for data collection, storage, and retention to prevent unauthorized access or misuse of personal information.
The decision, finally, to incorporate AI in P/CVE efforts should be undertaken deliberately and with a thorough grasp of its benefits and hazards. While AI has the potential to increase the efficacy of CSOs, it should be viewed as a tool to supplement rather than replace human decision-making. P/CVE practitioners’ skills and contextual knowledge are crucial in deciphering AI-generated insights and making informed decisions.
Kevin Waigwa is Head of ICT and Administration, Epuka Ugaidi Organization