As represented by the Moore's Law, the information technology up to this day has intensively pursued the "improvement of performance" including speedup, capacity increase, integration increase, functional improvement and low power consumption, as a target for the R&D. As a result, information systems have penetrated deeply and intricately into the society, so called "Information Society" is created where every activity between human and organization is heavily relying on the networked information system. In the information society, it is unable to draw a clear borderline between the information systems and the social system, and the information technology and the social technology function together for the system development and implementation.

Considering the prospects of the future matured information society under these circumstances, the direction for the R&D on the information technology ? as a national science and technology policy ? should not be the conventional "pursuing improvement of performance" but be the "pursuing dependability." The "dependability" is a technological concept that serves the core part of the safety, credibility and security in the information society, which is the universal philosophy for the science and technology in plotting the grand design of the future information society.

In the information society, when an event contradict to the expectation and understanding of the society is once occurred, such as the system failure, serious infrastructure accident, cyber terrorism, information leakage, etc., it may bring in serious situations as property loss, fatal accidents, paralysis on social and economic systems, etc. Even worse, the situation may become the threat to the national security. The society we should aim at is a dependable information society where people and organizations can trust the quality of services (dependability and safety) provided by the social infrastructure and information environment in order to realize a safe-and-peaceful life and satisfactory activities relying on the quality services.

However, we are facing with various risk factors that may block the dependability of the information society including black box, complicity and hugeness of system, minuteness of VLSI, explosive increase of information volume, diversification of services, aging of system elements, and uncertainty of responsibility in the network. These factors will further increase in the future along with the advancement and development of the information society.

In order to realize the dependability of the information society and establish a permanent social safety, credibility, and security assuming the presence of these risk factors, it is necessary to realize the dependability and evaluation for each of the following four classes composing the information society:

1) Backbone networked "Information system";
2) Social "Important infrastructure" constructed by the utilization of information system;
3) "Services and information" provided by the utilization of important infrastructure; and
4) "Information society" created by the people and organization receiving the service and information

For that purpose, we propose a comprehensive and basic R&D that should be carried out strategically and permanently for each of the four classes in following aspects:

i. An architecture that assures dependability permanently;
ii. Design and maintenance technique assuming life cycle risks;
iii. Quantitative evaluation technique for dependability.

In addition, in order to promote the R&D strategically and intensively in long-term viewpoint, we propose to establish a R&D base for handling basic assignments on safety, credibility, and security of the information society.