To Counter Domestic Extremism, Human Rights First Launches Pyrra

New enterprise uses machine learning to detect extremism across online platforms

WASHINGTON – Today, Human Rights First announced a new enterprise, originally conceived in its Innovation Lab as Extremist Explorer, that will help to track online extremism as the threats of domestic terrorism continue to grow.

Human Rights First originally developed Extremist Explorer to monitor and challenge violent domestic actors who threaten all our human rights. To generate the level of investment needed to quickly scale up this tool, the organization launched it as a venture-backed enterprise called Pyrra Technologies.

“There is an extremist epidemic online that leads to radical violence,” said Human Rights First CEO Michael Breen. “In the 21st century, the misuse of technology by extremists is one of the greatest threats to human rights. We set up our Innovation Lab to discover, develop, and deploy new technology to both protect and promote human rights. Pyrra is the first tool the lab has launched.”

Pyrra’s custom AI sweeps sites to detect potentially dangerous content, extremist language, violent threats, and harmful disinformation across social media sites, chatrooms, and forums.

“We’re in the early innings of threats and disinformation emerging from a proliferating number of smaller social media platforms with neither the resources nor the will to remove violative content,” Welton Chang, founding CEO of Pyrra and former CTO at Human Rights First, said at the launch announcement. “Pyrra takes the machine learning suite we started building at Human Rights First, greatly expands on its capabilities and combines it with a sophisticated user interface and workflow to make the work of detecting violent threats and hate speech more efficient and effective.”

The Anti-Defamation League’s Center on Extremism has been an early user of the technology.

“To have a real impact, it’s not enough to react after an event happens, it’s not enough to know how extremists operate in online spaces, we must be able to see what’s next, to get ahead of extremism,” said Oren Segal, Vice President, Center on Extremism at the ADL. “That’s why it’s been so exciting for me and my team to see how this tool has evolved over time. We’ve seen the insights, and how they can lead to real-world impact in the fight against hate.”

“It really is about protecting communities and our inclusive democracy,” said Heidi Beirich, PhD, Chief Strategy Officer and Co-Founder, Global Project Against Hate and Extremism. “The amount of information has exploded, now we’re talking about massive networks and whole ecosystems – and the threats that are embedded in those places. The Holy Grail for people who work against extremism is to have an AI system that’s intuitive, easy to work with, that can help researchers track movements that are hiding out in the dark reaches of the internet. And that’s what Pyrra does.”

Moving forward, Human Rights First will continue to partner with Pyrra to monitor extremism while building more tools to confront human rights abuses.

Kristofer Goldsmith, Advisor on Veterans Affairs and Extremism, Human Rights First and the CEO of Sparverius, researches extremism. “We have to spend days and days and days of our lives in the worst places on the internet to get extremists’ context. But we’re at a point now where we cannot monitor all of these platforms at once. The AI powering Pyrra can,” he said.

Pyrra’s users, including human rights defenders, journalists, and pro-democracy organizations can benefit from using the tool as well as additional tools to monitor extremism that are coming from Human Rights First’s Innovation Lab.

“This is a great step for the Innovation Lab,” said Goldsmith. “We’ve got many other projects like Pyrra that we hope to be launching that we expect to have real-world impact in stopping real-world violent extremism.”