Anthropic Twitter/X Compromise Venti Pump And Dump
Notice: This page is a freshly imported case study from an original repository. While the original content had a similar format, some sections may not have been fully completed. Please help fill in any empty sections or any missing information you can find. If you are new here, please read General Tutorial on Wikis or Anatomy of a Case Study for help getting started.
Notice: This page contains sources which are not attributed to any text. The unattributed sources follow the initial description. Please assist by visiting each source, reviewing the content, and placing that reference next to any text it can be used to support. Feel free to add any information that you come across which isn't present already. Sources which don't contain any relevant information can be removed. Broken links can be replaced with versions from the Internet Archive. See General Tutorial on Wikis, Anatomy of a Case Study, and/or Citing Your Sources Guide for additional information. Thanks for your help!
Anthropic, an AI research and safety company based in San Francisco, is focused on developing reliable and beneficial AI systems with safety as a priority. Their team, with expertise across machine learning, physics, policy, and product design, creates AI models such as Claude, which are designed to be aligned, harmless, and secure. Recently, Anthropic's official X (formerly Twitter) account was compromised, and the hacker used it to promote a fake token called CLAUDE and an AI bot called Venti. The hack resulted in a rapid token pump, netting the attackers around $100K before the tweet was deleted. While Venti gained brief attention, it did not lead to lasting success or growth for the token, which has since seen a decline in value. Anthropic confirmed that no systems or services were compromised, and they are working with X to investigate the situation further. Anthropic later announced that the root cause of the breach has been identified.[1][2][3][4][5][6][7][8][9][10][11][12][13]
About Anthropic
Anthropic is an AI research and safety company based in San Francisco, focused on developing reliable, beneficial AI systems with safety at the forefront. The company’s interdisciplinary team, with expertise in machine learning (ML), physics, policy, and product design, conducts research and creates AI models that prioritize alignment and harmlessness. One of their flagship products is Claude, a series of advanced AI models, with Claude 3.5 Sonnet being the latest release. Anthropic's work involves building AI systems that can be used in a variety of applications, from custom experiences to enterprise solutions, while also ensuring that the technology adheres to strict safety protocols. Through their AI research and innovations, Anthropic aims to address critical issues related to AI alignment, fostering systems that are both useful and secure.
The Reality
This sections is included if a case involved deception or information that was unknown at the time. Examples include:
- When the service was actually started (if different than the "official story").
- Who actually ran a service and their own personal history.
- How the service was structured behind the scenes. (For example, there was no "trading bot".)
- Details of what audits reported and how vulnerabilities were missed during auditing.
What Happened
The specific events of the loss and how it came about. What actually happened to cause the loss and some of the events leading up to it.
| Date | Event | Description |
|---|---|---|
| December 17th, 2024 8:39:00 AM MST | Venti Token Post Retweeted | The official Anthropic Twitter/X account retweets a post about the $Venti project, which has a token. |
| December 17th, 2024 9:04:00 AM MST | Start Of Breach Time | The start of the breach, according to the Anthropic AI History Status page. They report on their official status page at this time, that an unauthorized post has already been removed, suggesting that the post was actually made prior to this time. |
| December 17th, 2024 9:34:00 AM MST | Another Token Launch Post | A second post is made announcing a new $CLAUDE token which "will help accelerate, promote, and incentivize innovation in the crossroads of the cryptocurrency and AI space." |
| December 17th, 2024 2:16:00 PM MST | End Of Breach Time | The end of the breach, according to the Anthropic AI History Status page. |
| December 17th, 2024 2:57:00 PM MST | Anthropic Update Posted | Anthropic announces that the root cause of the breach has been identified. They have recovered control of the Twitter/X account. |
Technical Details
"The hacker shared both VENTI and a token pretending to be official with the ticker CLAUDE. One of the coins linked to the hack pumped immediately, leaving the exploiters with around $100K after a rapid rug pull."
Total Amount Lost
"The hacker shared both VENTI and a token pretending to be official with the ticker CLAUDE. One of the coins linked to the hack pumped immediately, leaving the exploiters with around $100K after a rapid rug pull."
The total amount lost has been estimated at $100,000 USD.
Immediate Reactions
The exploit lasted for about 30 minutes, during which the hacker shared the Venti bot, which was trained to communicate in the style of social media and marketed with irreverent internet jargon. The token associated with the hack, CLAUDE, experienced an immediate pump, allowing the attackers to profit around $100K before the tweet was deleted.
"The official X account of AI startup Anthropic, backed by Amazon, appears to have been compromised, posting an unknown token contract address related to AI Agents."
Ultimate Outcome
"We have identified the root cause which resulted in unauthorized posts on this account today.
We have confirmed that no Anthropic systems or services were compromised or involved in this incident. We're working with @X to better understand this situation."
While the promotion helped Venti gain brief attention, it failed to translate into significant success or following for the token, which has since been declining in value.
Total Amount Recovered
There do not appear to have been any funds recovered in this case.
What funds were recovered? What funds were reimbursed for those affected users?
Ongoing Developments
What parts of this case are still remaining to be concluded?
Individual Prevention Policies
No specific policies for individual prevention have yet been identified in this case.
For the full list of how to protect your funds as an individual, check our Prevention Policies for Individuals guide.
Platform Prevention Policies
Policies for platforms to take to prevent this situation have not yet been selected in this case.
For the full list of how to protect your funds as a financial service, check our Prevention Policies for Platforms guide.
Regulatory Prevention Policies
No specific regulatory policies have yet been identified in this case.
For the full list of regulatory policies that can prevent loss, check our Prevention Policies for Regulators guide.
References
- ↑ @AnthropicAI Twitter (Accessed Jan 27, 2025)
- ↑ Home \ Anthropic (Accessed Jan 27, 2025)
- ↑ Anthropic’s X account got hacked. - The Verge (Accessed Jan 27, 2025)
- ↑ Anthropic Status - Incident History (Accessed Jan 27, 2025)
- ↑ @btibor91 Twitter (Accessed Jan 27, 2025)
- ↑ @IaimforGOAT Twitter (Accessed Jan 27, 2025)
- ↑ @btibor91 Twitter (Accessed Jan 27, 2025)
- ↑ @savngmy_sol Twitter (Accessed Jan 27, 2025)
- ↑ @qtnx_ Twitter (Accessed Jan 27, 2025)
- ↑ Anthropic X account was compromised to promote a crypto AI agent (Accessed Jan 27, 2025)
- ↑ https://old.reddit.com/r/singularity/comments/1hggj8j/first_it_was_openai_now_anthropics_account_got/ (Accessed Jan 27, 2025)
- ↑ CLAUDE Token Scam: Anthropic's X Account Hacked for $100,000 (Accessed Jan 27, 2025)
- ↑ @AnthropicAI Twitter (Accessed Jan 27, 2025)