Spent nuclear fuel (SNF) is an intriguing problem from a safeguard instrumentation perspective: for safety reasons, it must be packed with a large amount of shielding to prevent radiation exposure to facility workers and the public, and with neutron-absorbing materials to prevent criticality. SNF is typically also at high temperature in dry storage and during transportation, precluding the use of many types of radiation sensors internal to a fuel cask. The radiation environment of SNF casks includes intense gamma radiation originating mostly from fission fragments, and a smaller but still sizable amount of fast neutron radiation, both from spontaneous fission of transuranic isotopes and (alpha,n) reactions with oxygen in the fuel matrix. Fast neutrons emitted by SNF are of particular interest for safeguards, as plutonium is the most significant contributor to the neutron source term. A gamma "compensated" He-4 ion chamber is in some ways an ideal instrument for use in SNF safeguards: He-4 has a high cross-section for fast neutron elastic scattering, it does not contain any materials that are consumed or transmuted, and it is a simple gas-based detector that is resilient to high temperature and radiation. The detector must be primarily sensitive to fast neutrons, as the thermal neutron population in a cask is depressed by neutron-absorbing materials.The gamma-ray contribution to the chamber current is compensated by sheathing two identical tubes in different thicknesses of tungsten, a strong attenuator of gamma rays. In this way, the fractional difference in current between the two tubes is directly proportional to the fractional contribution of gamma rays to the signal in the thinner tube, which can then be subtracted to give the fast neutron-only signal. In this work, a prototype detector system is evaluated using Monte Carlo simulations of a realistic SNF cask and fuel diversion scenario, and the gamma-compensation relationship is validated across the lifetime of a typical SNF cask.