The Center on Terrorism, Extremism, and Counterterrorism (CTEC) announced today that it has partnered with Spectrum Labs, a San Francisco-based startup that develops cutting-edge AI for the detection of hateful and violent content on the Internet — in context and across multiple languages and industries. CTEC will provide its expertise in extremist messaging to assist Spectrum in expanding their capacity for detecting coded words and idioms in non-English languages, starting with Portuguese and Spanish.
Extremism and terrorism are not U.S.-centric; hateful and violent speech is a problem in virtually every major language. CTEC has tracked white supremacy, racism, and violent extremism across cultures, developing an understanding of where it converges and diverges with English-language extremist speech. In partnership with Spectrum, CTEC will build out a knowledge base for monitoring such speech in Spanish and Portuguese, while remaining sensitive to cultural context and localized trends.
CTEC has established a diverse team of researchers and advisors. The team includes:
- MA students from the International Policy and Development, Translation and Interpretation, and Nonproliferation and Terrorism Studies programs;
- a Middlebury College student in the International and Global Studies program;
- faculty advisors from the Graduate School of Translation, Interpretation, and Language Education.
CTEC experts oversee this team and lead research into Brazilian and Latin American extremist movements and trends.
Spectrum Labs is producing innovative technology that aims to more safely, sensitively, and contextually detect and enforce against harmful speech, something that CTEC frequently calls for in our reports and recommendations. Its work has resulted in reductions of 75% or more in targeted toxic behaviors across social media, online marketplaces, dating, and gaming, among others. We are excited and honored for the opportunity to partner with Spectrum to work at the cutting edge of mitigating online extremism.