What happened
Replaying the Lynda.com October 2016 intrusion
- 1
Stolen credentials open GitHub accounts
October to December 2016, day not statedStolen user data opens several employees' GitHub accounts, and the accounts are searched for AWS credentials.
- 2
The AWS credentials are accepted
Before Dec 11, 2016, day not statedAWS accepts the credentials found through GitHub; their type, owner and permissions, and the network of the calls, are not stated.
- 3
Lynda.com user accounts downloaded
Before Dec 11, 2016, day not statedOver 90,000 confidential Lynda.com user accounts are downloaded from Lynda.com's AWS account.
- 4
Extortion through the bug bounty
Dec 11, 2016 to about Jan 16, 2017The intruders claim a security flaw to LinkedIn, send a data sample and are invited to its bug bounty program; the company pays nothing.
Over 90,000 confidential Lynda.com user accounts were downloaded from Lynda.com's AWS account, per the plea agreements. In December 2016 Lynda.com reset passwords for fewer than 55,000 users and notified about 9.5 million more whose learner data sat in the breached database, per The Register. The extortion failed: the company paid nothing.
Where cloud-provider controls fit
The GitHub accounts and whatever held the AWS credentials sat outside every cloud provider. The chain met AWS only when credentials found through GitHub were presented to the AWS API. On your clouds you can bind a credential's use to your own networks, with a service control policy on AWS, an IAM policy with a network source on Oracle Cloud and Conditional Access for workload identities on Microsoft Entra. You can refuse new long-lived keys, so nothing created afterwards can be left in a code hosting account. You can cap what a stolen credential reads in storage with a resource control policy, a permissions boundary and a KMS key policy. Over 90,000 confidential Lynda.com user accounts were downloaded before the intruders asked LinkedIn to pay.
Could this have been prevented?
In hindsight, no control would have stopped it. 9 controls would have limited how far it went. 3 steps have no cloud-provider control. The counts below are control assessments, not attack stages: several can address the same step. The same replay with the controls in place before the incident shows where each phase would have been refused or bounded.
- 0
- Prevented
- 9
- Contained
- 3
- Not covered
How it could have been prevented
The same replay with the controls in place. Green marks where a step is refused, blue where it is bounded, and red where no control reaches it. Open a row to read the control, why it acts, and its caveat.
Replaying the Lynda.com October 2016 intrusion
1 · Stolen credentials open GitHub accounts
2 Contained2 Not covered2 · The AWS credentials are accepted
3 Contained5 on other clouds3 · Lynda.com user accounts downloaded
4 Contained3 on other clouds4 · Extortion through the bug bounty
1 Not coveredKey facts
- From about October 2016 to about January 2017, the intruders used stolen credentials to obtain companies' data from their AWS accounts, per their plea agreements
- They tested stolen user data as GitHub credentials with a custom-built checker and searched several employees' GitHub accounts for AWS credentials; the pleas state this for the conspiracy as a whole, not for Lynda.com by name
- Over 90,000 confidential Lynda.com user accounts were accessed and downloaded from Lynda.com's AWS account before December 11, 2016; the day is not stated
- On December 11, 2016 the intruders emailed LinkedIn's security team claiming a security flaw and sent a data sample; LinkedIn invited them to its bug bounty program and paid nothing
- In December 2016 Lynda.com reset passwords for fewer than 55,000 users and notified about 9.5 million more, per The Register
- Both intruders' plea agreements were filed on October 30, 2019
Sources
Every fact in this breakdown comes from the public reporting below. The replay is a model of that reporting, not a reconstruction of the environment it describes.
- 1.Plea Agreement, United States v. Brandon Glover, No. CR 18-00348 LHK (Document 48)US District Court, Northern District of California (plea agreement filed by the US Attorney's Office; copy hosted by ISMG) · Oct 30, 2019 · AdvisoryThe conspiracy's dates, stolen user data tested as GitHub credentials, employees' GitHub accounts searched for AWS credentials, the credentials used against Amazon S3, the Lynda.com download from its AWS account and the count, the extortion contact, the data sample, the bug bounty invitation and no payment
- 2.Plea Agreement, United States v. Vasile Mereacre, No. CR 18-00348 LHK (Document 50)US District Court, Northern District of California (plea agreement filed by the US Attorney's Office; copy hosted by ISMG) · Oct 30, 2019 · AdvisoryThe same method and the same Lynda.com paragraph as the first plea: the GitHub path, the AWS credentials used against Amazon S3, the Lynda.com download and count, the extortion contact
- 3.Two Uber Hackers Plead GuiltyGovInfoSecurity (ISMG) · Oct 31, 2019 · NewsThe pleas, Amazon S3 as the store of the Lynda.com data, and Lynda.com's refusal to pay
- 4.LinkedIn's training arm resets 55,000 members' passwords, warns 9.5mThe Register · Dec 18, 2016 · NewsLynda.com's December 2016 notice: a database with learner data breached, fewer than 55,000 passwords reset and about 9.5 million users notified
The estate in the replay is illustrative and drawn at the level of attack stages; the victim's real environment is not public. Verdicts describe what a control does when it is in place before the incident. They are not an assessment of the victim's security program.