The oldest and truest cliché about data privacy is that technology is both the problem and the solution.

This is the cleft stick in which the government finds itself when it comes to the use of information in the growing number of databases storing our personal details.

There have already been murmurings that ministers want to relax elements of the Data Protection Act to allow further cross-matching of data in different Whitehall systems.

Where such a practice has been put in place, there have been successes. For example, cross-referencing between visa applications and the police fingerprint database has led to the arrest overseas of people that committed crimes in the UK.

Perhaps the highest-profile example is the car tax disc web site, which combines information on MOT results and motor insurance details to deliver one of the better online public services.

But the downside of this cross-matching has privacy campaigners up in arms.

What might the government learn about each of us were it to piece together fragments of our lives scattered across disparate departmental databases?

The fundamental principle of data protection legislation is that personal data should only be used for the purpose for which it was originally recorded. Excessive cross-matching would be a clear and controversial infringement ­ but you can be sure the government will look for more situations where it can claim that the benefits outweigh the risks.

The difficulty will come from the fact that trust in the way government uses our data is at an all-time low. HM Revenue & Customs has been told in no uncertain terms that it must develop a culture of information security and data protection as the price of the failings that led to the loss of CDs containing 25 million child benefit records.

The rest of government needs to learn the same lesson ­ as do corporations. The potential benefits of information sharing, cross-matching and data mining are real ­ but the risk can only be justified when the right culture and processes are in place.