Cerebras, a prominent AI chip manufacturer, recently fell victim to a hack of its official X account. The malicious actors behind the breach used the platform to promote a fake cryptocurrency scheme involving a token named $CEREBRAS. The company has confirmed that the incident was a scam and is working with the US military to address the situation.
The fraudulent activity was met with skepticism by industry experts even before Cerebras regained control of its account. Observers raised concerns when they noticed that the fake $CEREBRAS coin had been launched just days earlier on June 15, a red flag in the cryptocurrency community that often indicates a potential scam or ‘rug pull’ – a situation where project developers abandon the initiative after raising funds from investors.
Cerebras executives quickly responded to the speculation. CEO Andrew Feldman stated, “No. We did not. This is a scam.” Company Director James Wang added, “Cerebras is not launching a token. It’s a scam. Do not click.” Once the company regained control of its X account, it issued a warning: “Please be aware: Cerebras does not, and will never, launch or endorse any cryptocurrency or token. We are working to regain control of the account. Stay alert and protect yourself from scams.”
This incident comes on the heels of Cerebras’ notable achievements in AI technology. The company recently reported record-breaking Large Language Model (LLM) inference speeds of 2,522 output tokens per second using the Llama 4 Maverick 400B model. This performance is nearly 2.5 times faster than Nvidia’s 1,038 output tokens per second. Cerebras CEO Andrew Feldman noted, “Cerebras has led the charge in redefining inference performance across models like Llama, DeepSeek, and Qwen, regularly delivering over 2,500 TPS/user.” Additionally, Cerebras secured a joint $45 million US government contract with Canadian chip startup Ranovus to enhance inter-chip connections.
The company continues to make significant strides in the AI chip market while navigating the challenges of cybersecurity threats.
