The Snapchat breach that led to millions of user names and phone numbers posted on the Web shows how a startup's priorities of growth and features can sometimes lead to weak security, experts say.[CSO's guide to the Target data breach]While the Snapchat break-in was seen as avoidable, security experts had less to say about the hacking of Skype's official blog and social network accounts, saying not enough details were available. Both attacks occurred this week. With Snapchat, the mobile photo-sharing service made several errors in its use of cryptography and key management, according to Zak Dehlawi, senior security engineer for Security Innovation. Those mistakes led to the perpetrators posting 4.6 million user names and phone numbers on a Web site called SnapchatDB.info. The site has been taken down. On Christmas Eve, Gibson Security, based in New Zealand, posted a lengthy explanation of the vulnerabilities it found in Snapchat security after notifying the company of the problems. On Dec. 27, Snapchat said in its blog that it had implemented safeguards and counter-measures to prevent third-parties from matching user names and phone numbers and publishing them in a database, which is exactly what hackers did this week. In studying the Gibson findings, Dehlawi said the mistakes that led to the compromise reflect the difficulty developers, even seasoned ones, have in implementing cryptographic technology in mobile applications. "Any application that uses cryptography should undergo third-party code audits by security experts to identify security vulnerabilities," Dehlawi said."Second, companies should also ensure that their backend APIs (application programming interfaces) are secure by undergoing third-party security audits and following strict coding guidelines."The backend of applications often get shortchanged because developers assume it is not as vulnerable as the frontend, Dehlawi said.Wolfgang Kandek, chief technology officer for Qualys, said he was not surprised Snapchat was having trouble with securing its APIs. "I think this is almost normal for a company at their stage that is focused mainly on scalability and functionality," he said. "I am sure they will pay more attention to abuse and security issues in the future."[Alert: Your password is probably compromised...again]With Microsoft-owned Skype, the Syrian Electronic Army, a hacker activist group sympathetic with the Assad regime, hacked the Web-calling service's blog and posted the headline "Hacked by Syrian Electronic Army.. Stop Spying!" the Chicago Tribune reported.On the Skype Twitter account, the SEA posted contact info for retiring Chief Executive Steve Ballmer and the message, "You can thank Microsoft for monitoring your accounts\/emails using this details. #SEA," the newspaper said.SEA is notorious for defacing the Web sites and social media accounts of major media it views as unfavorable to President Bashar al-Assad in the Syrian civil war. The defacements appear to refer to revelations last year that Skype calls were monitored by the U.S. National Security Agency.Kandek said the compromise seemed to involve credential theft. If that's the case, then having two-factor authentication in place might have prevented the hackers from accessing the accounts on computers not recognized by the sites. Security breaches are inevitable for most organizations, so the strongest defense is to follow best practices and security standards relevant to the business, Larry Slobodzian, senior solutions engineer for LockPath, said. In addition, a data audit should be done, so security measures can be reinforced around the most sensitive information, and an up-to-date plan should be in place that describes how to respond when a system is hacked."It's a proactive approach before the breach happens," Slobodzian said.