Previous SectionIndexHome Page

Mr. Key: This is an important group of amendments and we have only a little over half an hour in which to have a coherent debate. I shall do my best to be speedy. The Minister is profoundly mistaken to think that the academic community will be in any sense reassured by what he has said and what his colleagues said in the other place. There is a misunderstanding about why the academic community is concerned. The current regime of export controls was introduced in 1939, and for many years it applied only to physical goods. Although technology could be transferred electronically by telegraph from the 1840s, by telephone and fax from the 1870s, and by radio from the 1900s, no attempt was made to do so.

7 pm

Some years ago the Department of Trade and Industry started applying export controls to manuals for weapons as well as to weapons. That caused a certain amount of grumbling, and served no obvious purpose other than to make work for the civil service. The DTI, however, has never lacked powers against genuine arms exporters. If a company successfully argued in court that controlling manuals was ultra vires, it would no doubt get hammered on its other licence applications. The problem is that these powers have often been used erratically. For example, the Canadian army was refused an export licence for flails for use in mine clearance in Bosnia. Apparently, the DTI saw the word "mine" and froze. That led to the Scott inquiry, after which something had to be done.

24 Jun 2002 : Column 666

During the 1990s, the United States National Security Agency tried to tighten the regulation of civilian cryptography. In the last months of the first Bush Administration, it pitched the clipper chip—a new standard for encryption with the feature that the National Security Agency would hold a master key. Thus US business could get the benefits of strong encryption, but if it were abused by criminals, the National Security Agency would be able to defeat it. George Bush senior and his Cabinet were hostile; they thought that it was bureaucratic empire-building and would cause them all sorts of trouble. Once President Clinton was elected, however, the pitch was repeated and got a warmer reception. Vice-President Al Gore took on the role of directing cryptography policy in the Cabinet. His efforts to sell the clipper chip were hampered by the availability world wide of encryption software that can be downloaded from many academic and other websites. Why should people pay $100 for encryption that the United States Government can break when they can get free software that they cannot break?

Cryptography is now used in a large number of systems as the glue that holds things together. It is found in every personal computer, in web servers, browsers, banks' automatic teller machines, car door openers, electricity meters and even one-armed bandits. As the USA tried to retain export controls on the technology, it waded deeper into trouble.

United States law, unlike English law, allows export controls to be applied to intangibles such as software downloaded from websites. There were furious battles in the United States about whether encryption software was the constitutionally protected academic speech of the professor who wrote it, to which the answer was "Yes, sort of", and whether it could be exported printed in a paper book, to which the answer was "Yes." People printed software on T-shirts that said, "This T-shirt is a munition." Companies such as Microsoft then poured money into civil liberties groups, annoyed that European encryption software firms were establishing markets close to them. Eventually, in the run-up to the 2000 election, Vice-President Gore caved in and removed most of the controls on the export of encryption software.

Before new Labour came to power, it adopted a strong position against the regulation of cryptography. Once it was in power, the line changed. The Government started putting out proposals for cryptoregulation that eventually, after much opposition and watering down, led to the Regulation of Investigatory Powers Act 2000. At the same time, consultation was started, and accelerated after the Scott report, on modernising the export control regime. A Green Paper came out in 1996 and a White Paper in 1998; the latter recommended the introduction of controls on transfer of technology by intangible means.

The best reference on this is the Trade and Industry Committee report of 10 December 1998 on strategic export controls. It took a critical view of the White Paper's line on intangible controls—paragraphs 37 to 42 are the key. The Committee found that the Government had no clear idea of the extent of the problem that the provisions sought to address, that there were grave doubts about the practicality of the proposals and fears for their consequences, and that there would be significant economic consequences such as the increase in costs of post-sales software maintenance and the support of online instructions manuals.

24 Jun 2002 : Column 667

The Committee noted severe worries about the effects on the operation and standing of United Kingdom higher education. It recommended that intangible export controls, if introduced, should be limited to weapons of mass destruction, and that the Government should not agree to an extension of the European Union's dual-use regime, whose aim was to bring exports of dual-use goods outside the EU under harmonised control.

The Government's reaction was to encourage the EU to put through a dual regulation in 2000 which, on the face of it, extended controls on the export of dual-use goods from the EU to intangible transfers. The legal view taken at the time by the universities was that this regulation had no force in criminal law because it would take primary legislation to introduce a new criminal offence, namely, conveying information to foreigners without the permission of Government. DTI officials claimed that the regulations did indeed have criminal force, but when pressed for an explanation of which Act would be used to prosecute, they were unable to help.

None the less, the DTI went ahead and started implementing dual-use controls on the community of arms manufacturing companies that it regulates. For such a company, a direct legal challenge could have unpleasant consequences. The view of the Defence Manufacturers Association at the time was—and, for all I know, still is—that the regulations also applied to tens of thousands of other companies which were, for example, incorporating cryptography in all sorts of products and exporting them while blissfully unaware of the alleged new legal situation. The DTI did not educate the companies which were now supposedly subject to licensing.

The Bill gives the Government almost arbitrary powers to regulate intangible exports. Officials say that it will be used sparingly. However, the EU dual-use regulation, which they will at last be able to enforce, will itself prevent, for example, technology transfers relating to cryptography outside the EU. Other technologies on the dual-use list include a very large part of what the Ministry of Defence considers to be high tech and thus a great deal of what interests universities.

Officials have been telling two different stories to academia. To Universities UK, they said that the Bill will be used to take powers to license the teaching of certain subjects to certain students. In public—and in Parliament—Ministers denied that that was the case. Lord Sainsbury tried to finesse the issue by saying loudly and often that the licences would be needed not by the students, as the Opposition had said, but by the professors teaching them. That is pure casuistry, as the students' name would also appear on the licence. Universities UK, whose chief executive is Baroness Warwick, did not push the issue.

Ministers say that the needs of academia will be met by exemptions covering pure science and information being put in the public domain. Yet the current arrangements for research grants for the Engineering and Physical Sciences Research Council require applicants to describe their proposal's "relevance to beneficiaries". Thus a pure mathematician wanting to do work on elliptical curves will describe it as being highly relevant to cryptography, which is the key enabling technology for electronic commerce, thereby putting his head in the noose.

24 Jun 2002 : Column 668

One academic told me that in 1997 he worked with scientists in Norway and Israel to develop cryptographic software for a competition run by the United States Government to find a replacement for a 1970s-generation encryption standard. DTI officials told him that in future, once they had intangible export controls in place, such an exchange would require a licence. As a result, the research would not get done. The conditions likely to be attached to such licences will not be acceptable to many universities or individual academics. Yet now we see Ministers equivocating as to whether a licence would be needed. In a letter to Baroness Warwick, Lord Sainsbury says that

Yet that relates to commercial development and thus appears to exclude free software; it needs a formal agreement, which must be registered; it excludes Chinese nationals; it has an onerous record-keeping burden; and it excludes cryptanalysis. No really useful scientific work on ciphers can be done without cryptanalysis. It is like trying to design aeroplanes without either a wind tunnel or computational fluid dynamics codes with which to test the designs.

Ministers claim that placing information in the public domain will make it exempt. That means that if I were to invent a global positioning system jammer that rendered US smart munitions ineffective, I could put the design on my web page so that the Taliban could download it and build it.

However, if I were to give a talk at a closed conference—a NATO conference, for example—I should need to obtain an export licence, not only for the talk but for any slides that I wanted to e-mail in advance to the conference organisers. Such a provision would also interfere with academic-industry collaboration, as we would appear to need licences to attend progress meetings. Perhaps all the problems will be sorted out in time—but certainly not tonight or by the Bill.

Control on the export of goods mainly affects trade. Important though the freedom of movement of goods may be, it is easy to understand the prior claims of public policy on national security. As restrictions on trade threaten jobs, the resulting tension between the requirements of national security and of the economy tend naturally to produce the necessary checks and balances on the export control power.

Control over the export of ideas is radically different, however. Freedom of expression is a fundamental human right and gives way much less readily to the claims of the state. Moreover, the natural checks and balances that affect decisions on trade are lacking, as blocking a research project will not normally produce job losses or other evident short-term economic effects. The extension of control from the export of goods to the export of intangibles is thus a radical step with serious implications; it is not merely the closing of a tiresome, recently discovered loophole in existing controls.

Export control over goods applies straightforwardly to goods being exported; ideas are far harder to control. The difficulty of the attempt to impose control on the export of ideas leads to fears of evasion and thus to subsections 2(c) and 2(d) of clause 2 giving power to impose control

24 Jun 2002 : Column 669

over the transfer of ideas—not their export but their exchange inside the UK. It is true that the power is qualified by the words

That is no help at all, however, as Universities UK has plainly told the DTI.

First, if the power can be used whenever technology may be used abroad, it can be used in practically every case, because any communication or publication of information may lead to its being used abroad. Secondly, if the power is used to impose controls that apply where

leaving it to the publisher or communicator to judge at his or her peril whether the control applies, the risk will be too high for many people. It is certainly obvious that there is always reason to believe that technology taught to foreign students may be used outside the United Kingdom, and therefore a clear power is needed to impose controls and licensing requirements on such students.

Two exemptions are needed to secure academic freedom; one relating to the public domain, and the second relating to the, in essence, private processes of research before publication. The public domain exemption divides into two elements: first, material already in the public domain and, secondly, the publication of the results of research that puts that material into the public domain. There should be no power to impose control on either of those processes.

The Minister disclaims any intention of imposing such controls, and points to the exemptions in the existing controls. Exemptions in present controls exist by grace and favour of the Minister. It is essential that they should be entrenched in primary legislation. Academic research involves much collaboration and communication between researchers. It is not made public as the research proceeds, and much of it may never be published. We all know about the false starts and rough drafts on the back of an envelope that are, in essence, private; their communication would not be protected by a public domain exemption. If the preliminary results are shared with research students, that may take place in seminars that are not open to the public, and would thus be equally unprotected by a public domain exemption.

What is required is an exemption for communications made in the ordinary course of academic research and teaching. The Minister referred to an exemption in existing controls for "basic scientific research", but there are two flaws. The first is that such exemptions are far too important to be left to secondary legislation. The second is that "basic scientific research" is far too narrowly defined, by being limited to exclude work with practical application. The Minister's Department insists that research grant applicants explain their work's "relevance to beneficiaries", thus ensuring that applicants classify their work in such a way as to take it outside the "basic scientific research" exemption.

Next Section

IndexHome Page