High-profile Twitter Users Compromised

High-profile Twitter Users Compromised

Wednesday 15 July 2020 saw the compromise of multiple high-profile Twitter users, including cryptocurrency exchanges, famous individuals and organizations, with their accounts subsequently being abused to Tweet cryptocurrency giveaway scams

Twitter- Figure 1 - Cryptocurrency Giveaway Scam via a compromised account

 

 

 

 

 

 

 

 

 

Figure 1 - Cryptocurrency Giveaway Scam via a compromised account

In addition to Tweets directly including a Bitcoin cryptocurrency address, bc1qxy2kgdygjrsqtzq2n0yrf2493p83kkfjhx0wlh, Tweets containing a link to a now offline website cryptoforhealth[.]com, hosted on a domain registered on the day of the attacks, encouraged victims to make a Bitcoin payment that would be 'doubled' and returned (Figure 2).

Untitled (30)

Figure 2 - Scam domain 'cryptoforhealth[.]com'

As is to be expected in scams of this nature, victims sending Bitcoin to this BTC address will not receive anything in return, reinforcing the age-old adage, "if something is too good to be true, it probably is".

'CryptoForHealth'

Whilst the Whois record for the cryptoforhealth[.]com domain is now 'redacted for privacy', it was initially registered, seemingly with false registrant data, to an 'Anthony Elias':

Registrant Name: anthony elias
Registrant Organization:
Registrant Street: 7528 Logan Street
Registrant City: Vacaville
Registrant State/Province: CA
Registrant Postal Code: 95687
Registrant Country: US
Registrant Phone: +1.3185553332
Registrant Phone Ext:
Registrant Fax:
Registrant Fax Ext:
Registrant Email: mkeyworth5@gmail.com

Passive DNS indicates that this site was hosted by Cloudflare Inc. (US) with multiple IP addresses, likely part of their content delivery network (CDN), resolving to the domain throughout the day:

  • 104.27.162.136
  • 104.31.238.10
  • 172.64.200.24
  • 172.64.201.24
  • 172.67.192.34

Note: Given that these IP addresses are part of Cloudflare's hosting infrastructure, network activity involving these is not necessarily indicative of a host visiting the scam domain.

Unverified reports also suggest that this domain was being delivered via messaging services (Figure 3), unrelated to Twitter, potentially indicating broader scam activity.

Twitter-Blog- Potential scam message (Via Twitter @ColePetersen14)

Figure 3 - Potential scam message (Via Twitter @ColePetersen14)

Further utilizing the 'Crypto For Health' moniker, an Instagram account seemingly takes responsibility for the incident with a subtle "It was us 🙂" message along with a stylized Twitter logo (Figure 4).

Twitter-Blog - 'Crypto For Health' Instagram account

Figure 4 - 'Crypto For Health' Instagram account

 

Compromise

Based on Tweets posted to Twitter's official support account, itself subject to abuse during this incident, Twitter are conducting an ongoing internal investigation. They believe that the accounts were compromised following a coordinated social engineering attack against Twitter employees to gain access to internal systems and tools.

Given this, it's assumed that these systems and tools were abused to gain control of the high-profile accounts and send scam Tweets on their behalf.

Notably, posts made to 4chan, an imageboard website, include screenshots allegedly showing Twitter backend interfaces (Figure 5) although their age and authenticity cannot be determined.

Twitter-Blog - Alleged Twitter backend screenshots (4chan /g/ - Technology)

Figure 5 - Alleged Twitter backend screenshots (4chan /g/ - Technology)

Further suggesting that these may be old screenshots, and somewhat unrelated to today's incident, the 'follower' and 'following' statistics do not appear to align with the three identifiable Twitter users, one of which currently has a display name and profile picture that is inconsistent with the 'backend' (Figure 6).

Twitter-Blog - Twitter profiles identifiable from the 'backend' screenshot

Figure 6 - Twitter profiles identifiable from the 'backend' screenshot

Whilst the full extent of this compromise is yet to be determined, with Twitter continuing their investigations and likely to provide updates via their @TwitterSupport [2] account, there is speculation that this could be an insider threat.

Again, whilst uncorroborated, an anonymous post on the 4chan website may provide an insight into Twitter's investigation (Figure 7).

Twitter-Blog-Alleged 'rogue employee' (4chan /g/ - Technology)

Figure 7 - Alleged 'rogue employee' (4chan /g/ - Technology)

Whilst Twitter's initial response saw them disable the ability for all verified accounts to Tweet, this capability was later restored at 2040hrs EST on Wednesday 15 July.

Consistent with a Twitter backend compromise, and as would be expected from high-profile accounts of this stature, numerous reports confirm that multi-factor authentication was in use by many of the accounts and therefore discounts any theory of individual account takeover.

 

IMPACT

Based on public reports thus far, the following notable accounts appear to have been abused/compromised:

  • AngeloBTC
  • Apple
  • Barack Obama
  • Bill Gates
  • Binance
  • Bitcoin
  • Bitfinex
  • Cash App
  • Charlie Lee (Litecoin)
  • Coinbase
  • Coindesk
  • Crypto Bitlord
  • CZ_Binance
  • Elon Musk
  • Floyd Mayweather
  • Gate.io
  • Gemini
  • Jeff Bezos
  • Joe Biden
  • Justin Sun
  • Kanye West
  • KuCoin
  • Mike Bloomberg
  • Ripple
  • TRON Foundation
  • Twitter Support
  • Uber
  • Warren Buffett

Based on the transactions associated with the threat actor's Bitcoin address, some 376 transactions totalling some 12.86BTC [1] were completed between 1900hrs UTC on Wednesday 15 July 2020 and 0805hrs UTC on Thursday 16 July 2020. Current exchange rates suggest that this would equate to US$117,385.

Given the high-number of followers and the social media influence these accounts hold, many of them have been previously impersonated and used in past cryptocurrency giveaway scams.

 

closing comments 

The question remains as to why a compromise of this apparent scale would be used to launch a relatively low yield cryptocurrency scam and utilize modus operandi that are somewhat more consistent with lower-sophistication threat actors. That being said, these attack motivations may align with the capabilities of an insider that was either complicit in providing access for a third-party, or directly abusing accounts themselves.

Conversely, this same apparent backend access in the hands of a higher-sophistication threat actor could presumably have been used to facilitate intelligence gathering operations against a number of notable individuals or even an attempt at misdirection to hide some other more nefarious activity.

Aside from this access potentially allowing private direct messages to be viewed, leading to potential abuse such as in extortion attempts, a motivated threat actor could have attempted to manipulate an account's followers by posting subtle Tweets. For example a government official's account could be abused to influence politics or a prominent business person's account might be abused to make announcements that influence stock market prices.


Regardless of the threat actor's motivation and intention, Twitter will likely need to consider how they limit and control employee access to user accounts and take measures to prevent similar incidents from reoccurring.

 

recommendations 

Although the true extent of this compromise remains unknown until Twitter concludes their investigation, many elements can already serve as a reminder to organizations when considering their own threats and risks.

  • Insider threats remain a threat to organizations of all sizes, especially as many employees may be working remotely at this time and are therefore without oversight. Implementing the principle of least privilege (POLP) should enable employees to access what they require for their duties whilst preventing access above this requirement. Additionally, measure such as requesting approval for additional access as well as robust logging and/or monitoring may dissuade employees from performing nefarious activities whilst providing damage limitation in the event of an employee account being compromised
  • Social engineering is a consistent tried and tested attack technique that can only be countered by educating users and empowering them to question suspicious activity. Aside from reminding users of the risks associated with unsolicited emails and conducting phishing exercises, consideration should be given to training high-risk individuals on other social engineer attack techniques such as threat actors posing as trusted parties in telephone calls.
  • Whilst the compromised accounts are not believed to be the result of account takeover, incidents such as this are a good opportunity to take stock of any organizational social media account and ensure that the credentials are adequately secured with multi-factor authentication where possible. Furthermore, where accounts are shared among a team, the use of password management tools may reduce the risk of compromise, whilst providing an audit trail.
  • Consideration should also be given to social media monitoring services that can alert an organization to any unauthorized or suspicious activity, allowing control to be regained, and limiting any brand or reputation damage.

[1]https://www.blockchain.com/btc/address/bc1qxy2kgdygjrsqtzq2n0yrf2493p83kkfjhx0wlh

[2] https://twitter.com/TwitterSupport/status/1283518038445223936


Next article