General Military Intelligence (GMI) Production. More attention should be devoted to integrating intelligence about operational geography and civil factors pertinent to military operations into over-all estimates. o Scientific & Technical Intelligence (S&TI) Production. Our Service planners and programmers would benefit from expanded analysis of S&TI function to include Third World arms production programs, weapons sales and thefts, and technology transfer. We look for improved integration of both HUMINT reporting and annotated imagery into S&TI production. S&TI databases on the Third World appear lacking. o Indications & Warning (I&W). Many non-military crises require a commitment of military resources for stability or humanitarian reasons. We are concerned by the absence of an estimative methodology and dedicated resources for anticipating such crises. Our community must have a "peacetime engagement" indications & warning capability, together with a capability to produce estimates relevant to national security planning and programming for Third World stability operations. o Human Intelligence (HUMINT). There are several general trends of concern: the most fundamental is that the existing national intelligence capability is simply not able to meet our needs for military and non-military plans and intentions; nor can it provide for contingency support, and stay-behind ground reconnaissance and support assets. This is especially the case in the Third World. o Signals Intelligence (SIGINT). The proliferation of commercial technology, the reduction of our overseas basing infrastructure, and the rapid emergency of multiple threat groups in new areas of concern (e.g. criminal and narco-revolutionary splinter groups in
areas of the world not previously covered) will make it extremely difficult for the SIGINT community to realign its resources and develop new capabilities with the declining dollars it receives under the defense draw-down. The SIGINT community is beset by other challenges, including a lack of qualified linguists for many lower priority languages. o
