Data integrity is a term used to refer to the accuracy and reliability of data. Data must be complete, with no variations or compromises from the original, to be considered reliable and accurate. Compromises to data integrity can happen in a number of ways. In industries where data is handled, identifying and addressing potential sources of damage to data is an important aspect of data security.
Problems with data integrity can start with a human source. People entering records may make mistakes, leading to variations between the original data and the data stored in a system. Likewise, people can make mistakes while transferring or copying data electronically, causing a disparity between different versions of a file or references to a file. In order for data integrity to be maintained, there need to be no changes or alterations to the data.
As data is entered, stored, accessed, moved, and updated, weak points in a system can compromise the data. Glitches in a computer may lead to partial overwrites of data or other data errors. Viruses can be created to attack data integrity, some working quietly to damage data without betraying their presence. Interruptions in various operations can lead to problems, as can mechanical damage like exposure to magnets or physical damage caused by power outages and other events.
Designers of data architecture for everything from government databases to the file systems on personal computers must consider data integrity when working on such systems. They think about the way the system will be used, identify obvious potential threats, and develop methods for securing the system to protect the data. Failure to think ahead can result in catastrophic compromises and the potential inadvertent release of secure or sensitive data, an especially large concern with databases containing personal or identifiable information about individuals and institutions.
People working with databases may be provided with training in data integrity, including reminders to review data as it is entered, to regularly save and back up data, and to report any suspected compromises or questionable activities immediately. The more quickly a problem is recognized, the easier it will be to address. Support personnel, such as information technology staff, also work on protecting databases from outside attacks with tools like firewalls, antivirus software, and periodic scans for malicious code. Likewise, systems can also be designed to lock data in read-only format to protect it from tampering or interference that might compromise its integrity.