[en] The revival of the civil nuclear power industry rekindles the debate on proliferation, in other words, on the increase in the number of States likely to achieve nuclear weapon capability. The history of the past thirty years shows that none of the known attempts at proliferation has occurred within the scope of civil nuclear program development. No country that has decided to 'proliferate' has done so by diverting materials or installations governed by the commitment to peaceful utilization and under IAEA control. The only borderline case is India, which did not sign the NPT, and which in 1974 cleverly played on the clauses imposed on it by Canada without violating them in the strictest sense of the word. The exporters of civil nuclear technologies subsequently got organized to control the export of sensitive materials by creating the Club of London, since renamed the Nuclear Suppliers Group. Another crisis occurred in 1991 following the first Gulf war, with the discovery of an extensive covert nuclear program in Iraq (an NPT signatory country); this caused a shock similar to that of 1974 and resulted in the strengthening of the IAEA's powers and inspection resources. India subsequently carried out several nuclear tests in 1998, at least one of which was a thermonuclear device. Pakistan, India's rival since the 1948 partition, crossed the 'nuclear threshold' in 1999. Proliferation also made headlines in 2003. First of all with Libya which, having decided to sign the NPT, revealed the existence of what is called the 'Nuke AQ Khan Bazaar'. In the presumed ignorance of his government, Dr. Abdul Qadeer Khan had stolen an uranium enrichment technology, becoming the 'father' of the Pakistani bomb, and had organized the traffic of military nuclear technologies, namely with Libya, North Korea and Iran. Again in 2003, North Korea announced its withdrawal from the NPT. This retreat raises the question of what will become of the possible transfers of nuclear technologies that could have been granted to this country under the Treaty. Even if proliferation has been more than just 'contained', there is always the risk that a State or a group of individuals will try to find illegal and covert channels in violation of international regulations. The decision to develop nuclear weapons is political and independent from the decision to install a nuclear reactor producing electric power. From a technological standpoint, two stages of the fuel cycle are 'sensitive' in relation to the risk of proliferation: uranium enrichment and the treatment of spent fuel. To obtain a quality fit for military use, the U-235 isotope must be enriched to a concentration of over 93%, but in practice this technique is only accessible to countries possessing the know-how. It is also difficult to use the plutonium derived directly from the spent fuel of light water reactors (generation II); moreover, the plutonium of spent MOX fuel is practically unusable for military purposes. Both these technologies are highly complex and are controlled by a restricted number of countries. Existing industrial enrichment and recycling facilities have nothing to do with military installations. They are of a limited number and are located in the five 'nuclear weapon States' (US, Russia, UK, China France), as well as in Japan, which has signed the IAEA Additional Protocol in complement to the full-scope safeguards agreement. The increase in the number of nuclear power plants around the world does not in any way signify an increase in the number of sensitive installations. Possessing one or several nuclear power plants does not mean that the entire industrial chain of the nuclear cycle also has to be developed to run them. Multi-lateral approaches to fuel supply are currently being investigated by the IAEA and its member States. The goal is to guarantee the supply of enrichment services over the long term to countries wishing to obtain nuclear reactors via multi-lateral mechanisms and commitments