01 — PRIMARY RISK AREAS
Artificial Intelligence
AI capabilities are developing at an exponential rate. AI experts are warning of potentially catastrophic consequences of the technology they themselves are developing, including misinformation campaigns, reduced barriers to bioterrorism, and rogue autonomous models.
READ FULL BRIEF →
Nuclear
Thousands of nuclear weapons around the world remain on hair-trigger alert — a false alarm or a single impulsive decision away from conflagration. Once launched, a missile cannot be recalled or shot down.
READ FULL BRIEF →02 — SECONDARY RISK AREAS
Biosecurity
Rising ability to engineer pathogens, and a history of laboratory leaks, threaten outbreaks of unprecedented danger. XLab's biosecurity research program is in early development.
Climate
Feedback effects beyond a tipping point could lead to catastrophic outcomes. XLab's climate risk research program is in early development.
More on These Risks
A House of Dynamite
With the Bulletin of the Atomic Scientists, XLab hosted a screening of Kathryn Bigelow's nuclear-war thriller.
The Doomsday Clock
A panel discussion on the Bulletin of the Atomic Scientists' annual Doomsday Clock announcement.
Rethinking Priorities for Frontier AI Security
A 2025 fellow's paper on prioritization in frontier AI security research.
Work on these problems with us.
Our Summer Research Fellowship trains the next generation of researchers across these risk areas.
EXPLORE THE FELLOWSHIP →