Anti-de Sitter Penrose inequality

A Penrose inequality asks how small the total mass can be given the size of a black-hole horizon. For initial data with one spherical anti-de Sitter end, can mass be bounded below by the area of the smallest cut enclosing the trapped boundary?

Why it matters and what counts as progress

Why it matters: The inequality connects horizon geometry with a global gravitational mass. Its equality case identifies which data attain the bound.

What counts as progress: Prove a sharp lower bound for a stated class of initial data, with explicit mass normalization, boundary conditions and area convention. A rigidity claim must describe the original data at equality.

How the problem developed

  1. 6 Oct 2026 · Reported nonmaximal inequality [1]

    OpenAI’s internal model The reported mass inequality m≥(r+r3)/2m\ge (r+r^{3})/2 holds for smooth three-dimensional exteriors with one spherical asymptotically anti-de Sitter end, dominant energy, the stated decay and a weakly future trapped compact boundary. The invariant mass must be future timelike, and r is computed from the infimum over all enclosing cuts. On the connected outermost outer-area-minimizing subclass, equality identifies the original data with a Schwarzschild–anti-de Sitter slice.

AI contribution

The nonmaximal anti-de Sitter Penrose Inequality and original-data rigidity

OpenAI

The reported mass inequality m≥(r+r3)/2m\ge (r+r^{3})/2 holds for smooth three-dimensional exteriors with one spherical asymptotically anti-de Sitter end, dominant energy, the stated decay and a weakly future trapped compact boundary. The invariant mass must be future timelike, and r is computed from the infimum over all enclosing cuts. On the connected outermost outer-area-minimizing subclass, equality identifies the original data with a Schwarzschild–anti-de Sitter slice.

AI involvementAI-led
ImportanceNotable
Technical verificationReported
Human understandingFew
Paper date:

Sources

  1. The nonmaximal anti-de Sitter Penrose Inequality and original-data rigidity
  2. OpenAI mathematics release
Publication history