The man behind DARPA's Memex now runs Microsoft's Catalyst Lab, and he is betting on cyber defense

On October 20, 2026, Microsoft Research published a podcast in which Weishung Liu interviews Chris White, vice president and managing director of the new Catalyst Lab. At DARPA, White led Memex, a program that helped prosecute human trafficking. The episode is mostly a profile, but it says something about how Microsoft wants to choose research problems.

From a Harvard postdoc to Afghanistan in one week

In the summer of 2010, White was a postdoc at Harvard. A mentor nudged him toward a DARPA meeting. The agency had pulled together social scientists, mathematicians, computer scientists and defense experts. For about 40 days they tested whether data analysis could help prevent violence and protect troops in Afghanistan.

At the end, DARPA asked whether he believed in his work. When he said yes, they sent him to prove it on the ground. He left within a week. The trip was meant to last nine days. On the third day he was put in charge, and he stayed 90 days, returning several times after that.

Later, as a DARPA program manager, he built Memex, a set of tools that connect the dots across hard-to-see parts of the internet. White says the tools helped New York City prosecutors bring ten times as many trafficking cases. In 2016 he received the Presidential Award for Extraordinary Efforts to Combat Trafficking in Persons.

Catalyst Lab brings DARPA-style problem selection into Microsoft

White calls problem selection the most important part of research. Microsoft Research, he says, offers a lot of freedom but expects ideas to travel all the way to proven value. His current priority is cyber defense: work on security, AI and hacking, protecting digital infrastructure and stopping attackers. The second direction is tools that help people analyze huge volumes of content. That is Memex in a new suit.

He describes the shift with his own image. At DARPA he worked with powerful, expensive lasers; at Microsoft he wants to make table knives. Both cut, and the world needs both. Liu also mentions that Catalyst Lab won a major commitment from leadership for a project roughly a decade in the making. Neither of them names it.

For security teams, this matters more than the life story. The DARPA model rests on a program manager who picks one hard problem, gathers several disciplines around it and demands proof in the field. If Microsoft really adopts it, the lab could ship defensive software rather than only papers.

The profile reveals more about the man than the lab

Much of the conversation covers scouting, yoga, a nickname from Kabul and Project Natick, the underwater datacenter built by White's predecessors. What Catalyst Lab is building today gets only a few general sentences. No tool, no result, no number.

The tenfold rise in New York prosecutions sounds impressive, but in the episode it is White's recollection without further detail. The other side of Memex is missing too. A tool that connects traces across the web helps prosecutors, and it serves surveillance just as well. The conversation does not address what safeguards similar work at Microsoft will have.

The first defensive tool Catalyst Lab ships will settle it

The first signal is whether Catalyst Lab publishes security research or a tool that people outside Microsoft can try. The second is the name of the project behind that leadership commitment.

With Memex, White could point to real court cases. Catalyst Lab deserves the same bar: deployment with real defenders and measurable impact, not just a good story on a podcast.

Lilith's verdict

Chris White wants to trade DARPA's expensive military laser for an ordinary table knife at Microsoft. How sharp that knife is will only become clear when network defenders use it to stop their first real attacker.