Impact Series

An engineer’s journey from preventing crime on the darkish internet to main ‘societal resilience’ at Microsoft

Chris White, general manager of Microsoft Research Special Projects, is leading a team that is developing open-source software tools and infrastructure to address crisis driven challenges. (John Edwards / TriFilm Photo)

Chris White believes in the power of technology to, if not save us, at least make the world a better place by helping safeguard the most vulnerable among us. He’s admittedly not alone among his engineering peers in holding an optimistic view of technology.

But what makes his position somewhat unusual is that White has confronted some of the ugliest, darkest corners of tech and emerged hopeful.

White is the general manager for special projects at Microsoft Research and part of a United Nations partnership that Thursday is releasing the first-ever global dataset of victims and survivors of human trafficking. The dataset was built using an open-sourced tool recently developed by White and others at the Redmond, Wash.-based company. It is being paired with a second tool that can help policy makers, law enforcement, nongovernmental organizations and others analyze the information to better understand and stop human trafficking.

The initiative is an example of the sort of mission-oriented, durable and collaborative tech solutions that White suggests are key to addressing complex crises. That includes human trafficking, as well as disasters that more directly touch all of our lives, like the COVID-19 pandemic and fallout from climate change.

“I’m looking toward the next crisis where we’re not waiting six months to understand whether it’s right to wear masks or not, whether we stay six feet apart or not. But that we’re prepared to be sharing information and putting evidence up for scrutiny in a way that is a bit more transparent,” White said. “And maybe that’s idealistic…but from my perspective, the best thing we can do is to inform the politics with the most accurate information.”

To that end, earlier this year White helped form a new organization within Microsoft Research group. It’s called the societal resilience team.

“We think it is critical to push this orientation of modern research to focus on global impact and empowerment, and to be relevant to the societies that we’re serving,” he said.

White began working at Microsoft in 2015. Before that, he was in Afghanistan as a program manager for DARPA (Defense Advanced Research Projects Agency). White led a program building software tools for digging into the dark web, prying into secretive places where information hides from conventional search engines. The project used machine learning to help make sense of what was unearthed. The technology was applied to sex trafficking rings, which by necessity use the dark web for conducting their business. The effort led to a surge in prosecutions and unraveled large trafficking networks.

Elisabeth, a woman who was trafficked at age 12, sits on a bed as she shares her story. (United Nation’s International Organization for Migration (IOM) / Lauriane Wolfe Photo)

DARPA, however, limits the tenure of its employees, so after four years White began looking for a new role. At Microsoft, he saw the potential to have a broader impact. Shortly after joining the software and cloud company, he helped form the Tech Against Trafficking coalition in 2018, with Microsoft acting as a founding member.

That work led to the collaboration with the UN’s International Organization for Migration (IOM) and creation of the Global Human Trafficking Synthetic Dataset. The information was gathered by IOM and major anti-trafficking organizations, and includes details gathered from case workers and hotlines.

Experts estimate that some 40 million people are trafficked, including children and people who are sexually exploited. The dataset captures information about 156,000 people from 189 countries.

“These data provide detailed insight into the profiles and experiences of the victims, the forms of human trafficking, and information on perpetrators,” said an IOM spokesperson by email. The dataset represents “a critical window into the crime and the hidden populations of victims and perpetrators, providing insights that other sources can’t.”

An image taken from the Global Human Trafficking Synthetic Dataset. (Click to enlarge)

As well as being the first of its kind, the public dataset is noteworthy for rigorously safeguarding the identities of the people included, while maintaining enough detail to provide essential information about trafficking. The researchers were able to do both by using an open-source tool called Synthetic Data Showcase that was built by Microsoft’s societal resilience team.

The Showcase tool creates synthetic records that cannot be traced back to individuals or groups of individuals, protecting their privacy and safety. The complete dataset in aggregate still contains accurate statistical information needed for analyses and decision making.

And that’s where the second open-source tool, ShowWhy, comes in. This software, which is also being announced Thursday and will be released later this year, helps reveal causal relationships from data. In this case it can reveal, for example, the impact of a natural disaster on trafficking or the effectiveness of public policy in curbing the crimes.

The IOM will oversee the maintenance of the trafficking dataset, while Microsoft Research will continue developing Showcase and ShowWhy.

White is eager to apply and adapt the tools to other challenges. He said that corruption would be a good area target, teasing out the roles of different players, tracking fraud and the misappropriation of resources. A strong dataset on corruption could ideally drive policy to combat it.

“We think,” said White, “that by combining data and technology and collective action, that will leave these activities nowhere to hide — not even the dark web.”

Editor’s note: This story has been updated to correct the release date of ShowWhy.

Related Articles

Trả lời

Email của bạn sẽ không được hiển thị công khai. Các trường bắt buộc được đánh dấu *