An article from CNet details the efforts of Intellect (an industry group made up of IBM, Microsoft, BAE Systems, Intel and others to get the UK government to steer clear of allowing GPL licenses for government funded projects.
It's pretty common for governments (including the UK and the US Federal government) to fund software development as part of grants for research. Generally speaking, much of this funded development ends up in some kind of Open Source license.
This is, it appears, not what the organization takes issue with. Instead, they are concerned about which kinds of Open Source licenses are used and, in particular, avoiding the GPL.
The particular concern is that the GPL has specifiic provisions that require all derived works must also be made available under the GPL, which requires that both the source and object code be available to anyone who wants it.
This may initially sound like a good idea for publicly funded projects, but in fact it is not. The reason is that the GPL effectively eliminates the commercial exploitation of government funded projects, while allowing the non- commercial exploitation. In particular, this means that the liklihood of seeing quality commercial software come out of government research is near zero.
I'm all for standardizing on a particular licensing form for federally funded projects, but my preference is that it be placed completely in the public domain, or that the license be more like the FreeBSD license.
The FreeBSD license requires that credit be given, but other than that doesn't make any requirements.
Now, I have a a bias here, because I've contributed software into the public domain and contributed to open source projects under the FreeBSD license. Further, InterCon Systems Corporation, the company that I helped run for 7 years started with a piece of software that I had written at NCSA (the same place that Mosaic came from) and which was made available in the public domain because it was federally funded.
At the time that we were selling TCP/Connect and TCP/Connect II (much enhanced versions of NCSA Telnet), we were competing with implementations from dozens of organizations in many countries, including Japanese localizations, Chinese localizations, and groups using the underlying code for completely different programs. Even today, there are still dozens of implementations of Telnet that trace their origins to that code written by two undergraduates and two graduate students at the University of Illinois from 1985-1988.
This is the power of releasing software to the public domain.