Terrorism early warning system to release new reports

• Reporters are invited to attend the release on Thursday
• Machine-learning platform models terrorist behavior to generate forecasts of future attacks
• Model can predict the likelihood of an attack by a specific terrorist group within the next six months
• Researchers have trained models for six terrorist groups: Abu Sayyaf, Boko Haram, Lashkar-e-Taiba, Indian Mujahideen, Al-Shabaab and Jama’at Nusrat al-Islam wal-Muslimin

EVANSTON, Ill. --- Tomorrow, Northwestern University artificial intelligence (AI) and security experts will release new reports from the Northwestern Terror Early Warning System (NTEWS), a machine-learning platform that models terrorist behavior to generate forecasts about future attacks.

The release will take place on Thursday, Oct. 17, at the Conference on AI & National Security. An introduction and system demo will occur between 9 and 9:45 a.m. CDT.

The Northwestern Security and Artificial Intelligence Lab (NSAIL) hosts the conference each year to explore new ways to integrate AI into national and global security strategies. In addition to the releasing the new reports, the conference will feature a range of presentations showcasing new AI technologies and panel discussions offering insights from leading researchers, security strategists and others.

Open to the public and external media, the day-long event will begin at 8:15 a.m. CDT at 1800 Sherman Ave., Suite 3-000 in Evanston. Registration is full, so members of the media must contact May Malone at may.malone@northwestern.edu to reserve a spot.

“The gradual drawdown of U.S. operations in Afghanistan and Iraq have led the world to a degree of complacency about terrorism,” said Northwestern’s V.S. Subrahmanian, who led the development of NTEWS. “Yet, the threat is omnipresent as we see continuing terror attacks in Africa from groups such as Al-Shebaab, Boko Haram and Jama’at Nusrat al-Islam wal-Muslimin (JNIM) as well as in Asia from groups such as Lashkar-e-Taiba, Abu Sayyaf and others. Counterterrorism organizations need the power to be able to predict the approximate time frames and types of attacks that terror groups might carry out in the coming months. Armed with such knowledge, they can better direct intelligence collection resources, allocate counter-terrorism resources and decide on optimal security strategies.”

An AI and security expert, Subrahmanian is the Walter P. Murphy Professor of Computer Science at Northwestern’s McCormick School of Engineering and a faculty fellow at the Buffett Institute for Global Affairs. He also is director of NSAIL.

By analyzing specific terrorist group activity, NTEWS aims to predict terrorist behavior in advance of any attacks they might carry out, allowing officials to plan for when attacks are likely to happen and mitigate their impact. These types of threats include attacks on security installations, security forces, public sites, individual people and more.

“NTEWS doesn’t predict attacks on specific locations or exact dates,” said Chiara Pulice, a senior researcher at Northwestern, who works on NTEWS. “But it can predict the likelihood of an attack by a specific group within the next month, two months, three months, four months, five months or six months.”

NTEWS leverages publicly available data from news outlets and other online sources about different types of attacks and various circumstances from the past several years. Researchers developed the first predictive models of Boko Haram and Lashkar-e-Taiba. Using those models, Subrahmanian and his team have made several policy recommendations over the years to reduce the number of Boko Haram and Lashkar-e-Taiba attacks.

Since then, NTEWS expanded its trained models to include predictions for Abu Sayyaf, Indian Mujahideen, Al-Shabaab and JNIM. Starting Thursday, NSAIL will begin releasing monthly NTEWS Terror Forecast reports on these six terrorist groups. To receive these reports, please fill out this form.

— May Malone contributed to this piece