Observations from the Atmospheric Infrared Radiance Sounder(AIRS) have been available since mid-2002. Of the high-volume 2378 channels, a carefully selected 281 set is processed for data assimilation and numerical weather prediction. The fifth generation of the Goddard Earth Observation system(GEOS-5), an operational data assimilation system currently being developed at the NASA Global Modeling and Assimilation Office(GMAO), employed a even more restrictive 118-channel set. This subset includes 84 long-wave temperature, 20 water-vapor and 14 short-wave temperature channels. Due to the overlapping of their weight functions, the use of some channels could have negative impacts on the use of other channels. A diagnostic tool is being developed to evaluate the channel interferences in terms of the analysis gain and information content. The information content is defined simply as the total error variance reduction due to the assimilation of the observations. In the case of AIRS observations, the information content can be partitioned into the components related to temperature, specific humidity and surface temperature due to each channel over each model level. The channel interferences could be reduced through the proper specification of the background and observational errors, so that AIRS observations, especially the observations from water-vapor and window channels, could be used better in the data assimilation.