How Twitter got hacked

by billso on Wednesday, 7 January 2009

When I wrote my billso.com arti­cle about the attacks on Twit­ter, I made some rea­son­able guesses about how the attacks might have been staged by using some phish­ing techniques.

While there was a spate of Twit­ter phish­ing attacks over the past week­end, it now appears that most, if not all, of the hijacked celebrity Twit­ter accounts were not phish­ing vic­tims. Twit­ter admit­ted in the Mon­day Morn­ing Mad­ness blog post that they were vic­tims of a dif­fer­ent kind of attack.

Wired’s Threat Level blog reported yes­ter­day that the ini­tial attack was actu­ally a dic­tio­nary attack on a Twit­ter employee’s account. That ini­tial attack was suc­cess­ful because:

  • Twitter’s login sys­tem kept pre­sent­ing a fresh login screen instead of cut­ting the user off after sev­eral incor­rect passwords
  • Crystal’s pass­word was a com­mon dic­tio­nary word: hap­pi­ness. The attack would have failed if Crys­tal had cho­sen a rea­son­ably strong pass­word or passphrase, or included one or two sym­bols or num­bers in the orig­i­nal pass­word. See my billso.com arti­cle from 24 April 2008 called Change that pass­word into a passphrase
  • Crystal’s Twit­ter account also had access to Twitter’s cor­po­rate account man­age­ment tools that let the hacker hijack any Twit­ter account they wished to use.

Wired’s arti­cle called Weak Pass­word Brings ‘Hap­pi­ness’ to Twit­ter Hacker has more details and a YouTube video that hacker GMZ offered to sup­port his claims.

You are the weak­est link”

Damon Cortesi pointed out in his 6 Jan­u­ary 2009 guest edi­to­r­ial called The inevitable rise (and fall?) of ‘twish­ing’, Twit­ter users seem more than will­ing to fol­low links sent through direct mes­sages (DMs) from their Twit­ter con­tacts. On 5 Jan­u­ary 2009, I pointed out a sim­i­lar issue in this billso.com arti­cle called Phish­ers hit Twit­ter. It’s a well-known flaw in Twitter.

Twit­ter has some self-regulating fea­tures built in to the ser­vice, and it’s easy for users to stop fol­low­ing con­tacts. Twit­ter doesn’t have any sort of spam fil­ter­ing, however.

Damon’s guest post appeared on Jen­nifer Leggio’s blog a week after Jen­nifer wrote that Twit­ter would get phished some­time soon. See her arti­cle called Friend­Feed, Twit­ter address URL redi­rec­tion risks; Face­book, LinkedIn, MySpace lag behind.

Jen­nifer also appeared on Jim Turner’s talk show yes­ter­day, along with Chris Messina, to dis­cuss Twitter’s bad week­end. See Social Media Secu­rity and the Twit­ter Phish­ing Trip to hear the audio record­ing and read the chat transcript.

At least one of the third par­ties that pro­vide Twitter-related ser­vices have started post­ing or explain­ing their pri­vacy poli­cies this week. Mr. Tweet has a rea­son­able expla­na­tion called Address­ing Pri­vacy Con­cerns. Over the next few days, I expect a few more sites to post their own explanations.

Related arti­cles and pages on billso.com

Share

Comments on this entry are closed.

Previous post:

Next post: