Web application security among UK corporates is getting worse, according to audits carried out by CESG-accredited security consultancy NTA Monitor.
NTA Monitor reports a ten percentage point increase in the total number of web applications found to have at least one high-risk security issue. A quarter (27 per cent) of all web applications tested by the penetration testing firm harboured at least one high risk issue, compared to 17 per cent in the previous year.
Incidents of high-risk security flaws vary widely across industry sectors, with utilities and legal firms doing particularly well at avoiding high-risk vulnerabilities. NTA’s service sector and not-for-profit clients, by contrast, fared badly.
The service sector had the highest number of high risks per test compared to all other sectors.
The latest edition of the annual Web Application Security report identified three types of high risk flaws, defined as vulnerabilities that might allow unauthorised external users to obtain system access, as listed below:
- SQL injection attacks, which allow attackers to modify the database queries initiated from an application.
- Cross-site scripting attacks, which creates a means for third-party websites to cause potentially malicious code to be run in a user's browser.
- Cross-request forgery attacks, a class of flaw that can allow a hostile website to make arbitrary HTTP requests to applications.
Roy Hills, technical director at NTA Monitor, explained: "All user-supplied data should be properly sanitised before returning it to the browser or storing it in a database. This reduces the threat of SQL injection, which is a consistently prevalent high risk throughout 2008 and 2009.
"SQL injection enables attackers to modify the database queries initiated from an application so users can delete, create or update database records."
NTA recommends three key steps that organisations can take to reduce security risks: sanitising user data before returning it to the browser or storing it in a database, switching from a persistent authentication method to a transient authentication method to help prevent cross-request forgery attacks, and applying an account lockout mechanism (to frustrate brute force attacks). ®