Key points

  • The Justice Department charged Hefu Chai and Huaisong Xiang with commodities fraud and wire fraud.
  • Prosecutors allege each defendant earned more than $50,000 by trading perpetual futures before Robinhood Crypto listing announcements.
  • The allegations are unproven, and both defendants are presumed innocent unless convicted.

Federal prosecutors charged two former Robinhood engineers with commodities fraud and wire fraud on September 15, alleging they used confidential information about upcoming cryptocurrency listings to trade perpetual futures on Hyperliquid. The Southern District of New York identified the defendants as Hefu Chai and Huaisong Xiang, also known as Jerry Xiang. The complaints are accusations, not findings of guilt, and both men are presumed innocent unless and until proven guilty.

Prosecutors describe trades before listing announcements

According to the Justice Department, Chai and Xiang had access through their engineering roles to nonpublic information about whether and when Robinhood Crypto would support additional tokens. Prosecutors allege they repeatedly bought perpetual futures tied to those tokens before Robinhood publicly announced the listings during 2025 and 2026. The agency said each defendant made more than $50,000 from the trades. A perpetual future is a derivative that tracks an underlying asset without an expiry date, allowing a trader to gain or lose from price movements without holding the token itself.

Related reporting: Senate blocks CLARITY Act in 49-50 procedural vote

The complaints outline different periods and access

The complaint against Chai says he served as a technical lead for new digital-asset listings and was among a restricted group with access to a private internal channel. It alleges he traded ahead of at least ten announcements between 2025 and January 2026, involving tokens including MEW, MOODENG, ASTER, XPL, HYPE and others. The broader Justice Department release says Xiang also traded using material nonpublic information. Those details remain allegations that prosecutors would need to prove in court.

Why the derivatives angle matters

The case extends the government's focus on confidential crypto-listing information beyond spot-token transactions. Prosecutors are relying in part on the Commodity Exchange Act because the alleged positions were perpetual futures on a derivatives venue. The Justice Department said each defendant faces one count under that law, carrying a maximum potential prison term of 10 years, and one wire-fraud count, carrying a maximum of 20 years. Maximum penalties are set by statute and do not predict an eventual sentence.

Robinhood says it reported the matter

Robinhood was not charged. In a statement reported by The Block and The Wall Street Journal, the company said it has zero tolerance for insider trading, investigated the conduct and reported the matter to law enforcement and regulators. The Justice Department also thanked Robinhood for cooperating. Chai was scheduled for an initial appearance in the Northern District of California, while Xiang was due to appear before a federal magistrate judge in New York, according to the official release.

What comes next

The cases will test how prosecutors apply established fraud and commodities rules to trading on decentralized crypto-derivatives platforms. They also highlight the compliance risks around token-listing calendars, which can be market-sensitive even when a platform does not publish advance notice. Employers with listing teams may face renewed pressure to restrict access, monitor conflicts and preserve audit trails. Market operators may also examine whether surveillance systems can connect internal access records with suspicious derivatives activity occurring before public announcements. That question is especially relevant when trading happens on venues outside an employer's own systems. For now, the central facts are that criminal complaints have been unsealed and the defendants have been charged; no conviction or final determination has occurred.

Sources

AI-generated editorial image; not a photograph of the reported event. Prepared with AI assistance and source verification.