FTX Halts Sale of $500 Million Stake in AI Startup Anthropic

Bankrupt cryptocurrency exchange FTX is putting the brakes on the sale of its stake in AI platform Anthropic, according to a Bloomberg report on Tuesday.

While no specific reason for the halt was given, FTX holds over $500 million worth of shares in Anthropic, according to a November 2022 report by the Financial Times.

FTX filed for Chapter 11 bankruptcy protection in November. A month later, FTX co-founder Sam Bankman-Fried was charged with several federal crimes, including money laundering, fraud, and conspiracy to commit wire.

In court documents filed yesterday in the United States Bankruptcy Court of Delaware, FTX’s former leadership, including Bankman-Fried, is accused of commingling over $402 million in customer funds at the direction of Bankman-Fried and other senior FTX executives. The failed exchange, the document said, owed customers approximately $8.7 billion, $6.4 billion in the form of fiat currency and stablecoins.

In January, a federal judge overseeing the FTX bankruptcy permitted FTX to sell off some of its assets to repay creditors.

Other assets that FTX has sold include derivatives trading platform LedgerX for $50 million, representing a considerable loss compared to the $300 million the crypto exchange paid for LedgerX in 2021.

Earlier this month, Semafor reported that FTX, through financial services company Perella Weinberg Partners, was looking to offload its shares in Anthropic.

The potential sale of the Anthropic shares was one of the attempts by FTX to “clawback” funds to pay off creditors.

A clawback in bankruptcy is a legal process where a bankruptcy trustee, in FTX’s case Perella Weinberg Partners, retrieves property or payments made by the company before filing for bankruptcy.

Launched in 2021 by former OpenAI employees, Anthropic launched Claude AI in March after receiving a $400 million investment from Google earlier this year, with another $450 million in Series C funding in May led by Spark Capital.

That same month, Anthropic announced advancements in Claude AI, including a set of rules based on the Universal Declaration of Human Rights, to promote ethical behavior.

This “constitution” sets specific training parameters to align Claude’s actions with ethical standards and deter undesirable behavior.

At a meeting at the White House with Vice President Kamala Harris, Anthropic, OpenAI, Google, Microsoft, Nvidia, Hugging Face, and Stability AI committed to an independent public evaluation of AI systems.

Perella Weinberg Partners declined to comment for this story.