We discuss some open problems concerning the origin and the emission mechan
ism of gamma-ray bursts (GRBs) in light of recent developments. If GRBs ori
ginate at extragalactic distances, we are facing four crises: (1) an energy
crisis, models have to account for more than 10(53) ergs of energy emitted
in the gamma-ray band; (2) a spectral crisis, emission models have to acco
unt for the surprising 'smoothness' of GRB broad-band spectra, with no indi
cation of the predicted spectral 'distorsions' caused by inverse Compton sc
attering in large radiation energy density media; (3) an afterglow crisis,
relativistic shock models have to explain the complexity of the afterglow b
ehavior, the longevity of optical transients detectable up to six months af
ter the burst, the erratic behavior of the radio emission, the lack of evid
ence for substantial beaming as indicated by recent searches for GRBs in th
e X-ray band; (4) a population crisis, from data clearly indicating that on
ly hard and long GRBs show a strong deviation from an Euclidean brightness
distribution, just the opposite of what expected from extragalactic models
without substantial cosmic source evolution. All previously proposed cosmol
ogical models are challenged: in particular, the neutron star-neutron star
coalescence model most likely will not survive the resolution of the proble
ms raised by points (1) and (4). INTEGRAL detections in the MeV range can g
reatly contribute in clarifying issues (2) and (4).