# United States v. Agrawal: When Printing the Code Made It a Crime

> The Second Circuit upheld EEA and NSPA convictions because a trader stole code on paper, distinguishing Aleynikov and exposing gaps Congress later closed.

Topic: Trade Secrets  |  Author: Lidiia Levitska  |  Source: Intellectual Property Law (outsideipcounsel.com)
Canonical: https://outsideipcounsel.com/blog/us-v-agrawal-paper-code-eea-nspa/


Two former Wall Street technologists stole proprietary high-frequency trading code from their employers within roughly a year of each other, took it to competitors, and were prosecuted under the same federal statutes. One conviction was reversed and one was affirmed. The difference was the medium. *United States v. Agrawal*, 726 F.3d 235 (2d Cir. 2013), decided August 1, 2013 in an opinion by Judge Reena Raggi of the United States Court of Appeals for the Second Circuit, upheld convictions under the Economic Espionage Act and the National Stolen Property Act because the defendant printed the code onto paper, distinguishing the court's earlier reversal in *United States v. Aleynikov*, 676 F.3d 71 (2d Cir. 2012), where the theft was purely electronic. Read together, the two cases exposed how much criminal liability could turn on physical form, and they help explain why Congress amended the EEA in 2012.

## At a glance

- **Case:** *United States v. Agrawal*, 726 F.3d 235 (2d Cir. 2013)
- **Decided:** August 1, 2013, majority opinion by Judge Reena Raggi, with Judge Rosemary Pooler dissenting in part; convictions affirmed
- **Holding:** Convictions under the EEA (18 U.S.C. 1832) and the National Stolen Property Act (18 U.S.C. 2314) stand where the defendant printed trading code onto paper, because the paper was a stolen tangible good and the code related to securities traded in interstate commerce.
- **Significance:** The essential companion to *Aleynikov*, showing that the medium of theft could decide criminal liability under the statutes as then written, and illustrating the gap Congress closed with the Theft of Trade Secrets Clarification Act of 2012.

## The parallel thefts

Samarth Agrawal worked as a quantitative analyst at the Paris-based bank Societe Generale, where he had access to the code underlying the bank's high-frequency trading systems. He secretly printed thousands of pages of that code and took the printouts to his home in New Jersey, intending to help replicate the systems for Tower Research Capital, a competing New York hedge fund that had offered him a lucrative position. He was convicted after a jury trial under both the EEA and the NSPA.

The prosecution unfolded in the shadow of *Aleynikov*. Sergey Aleynikov, a programmer at Goldman Sachs, had transferred the bank's high-frequency trading source code to an external server and then downloaded it to his own devices before leaving for a startup. He was convicted under the same two statutes, but in April 2012 the Second Circuit reversed. The court held that Goldman's trading system was not a product "produced for" or "placed in" interstate commerce within the meaning of the EEA as then written, and that the intangible transfer of source code was not the theft of "goods, wares, or merchandise" under the NSPA. Agrawal's appeal asked the same court to apply that reasoning to a defendant who had done something materially different with the code.

## Two statutes, one distinction

The National Stolen Property Act punishes the interstate transportation of stolen "goods, wares, merchandise, securities or money" worth 5,000 dollars or more. *Aleynikov* had held that purely electronic code, transmitted without any physical thing changing hands, fell outside that language; the defendant never assumed physical control over anything tangible. Agrawal's conduct differed at exactly this point. By printing the code onto thousands of sheets of paper and carrying them across state lines, he took physical control of a tangible object. The Second Circuit held that the printed pages were "goods, wares, or merchandise," so the NSPA reached the theft. The intangible became tangible the moment it was committed to paper, and that transformation carried legal weight.

The EEA analysis followed a related but distinct path. Section 1832, as then written, required that the trade secret be "related to or included in a product that is produced for or placed in interstate or foreign commerce." *Aleynikov* found that Goldman's internal trading system was not such a product, because the bank used it in house rather than selling it. In Agrawal, the court located the interstate-commerce hook not in the trading system itself but in what the system traded. The code was related to securities, and securities are unquestionably bought and sold in interstate commerce. That connection, the majority held, satisfied the statutory element even though the trading platform, like Goldman's, was never a product offered for sale. Judge Pooler dissented in part, disputing that the EEA element was met, a disagreement that foreshadowed the statutory revision to come.

## Why the medium mattered so much

The pairing of *Aleynikov* and *Agrawal* produced an uncomfortable result: two defendants who stole functionally equivalent assets faced opposite outcomes based on whether the theft passed through a printer. That the law could hinge on so arbitrary a feature was widely criticized, and the criticism was not merely academic. Trade secrets in the modern economy are overwhelmingly digital. A reading of the criminal statutes that protected paper but not electronic copies would leave the most valuable and most commonly stolen assets least protected, precisely inverting the statutes' purpose.

The Second Circuit did not pretend the distinction was elegant; it applied the statutes as Congress had written them. The court's fidelity to the text is itself the lesson. Where a criminal statute keys liability to tangible goods or to products produced for commerce, the government's case can rise or fall on facts that have nothing to do with the defendant's culpability and everything to do with drafting choices made before anyone imagined server-to-server code theft.

## The 2012 clarification amendments

Congress had already begun to respond. Reacting to the gap *Aleynikov* exposed, it passed the Theft of Trade Secrets Clarification Act of 2012, which amended section 1832 to cover a trade secret "related to a product or service used in or intended for use in interstate or foreign commerce," replacing the narrower "produced for" language. The revision was designed to ensure that internally used software, including trading systems like those at Goldman and Societe Generale, falls within the EEA regardless of whether it is ever sold and regardless of the medium in which it is taken. Because Agrawal's conduct predated the amendment, his case was litigated under the old text, which is why the securities-in-commerce theory had to carry the EEA count. The amendment means that a future Agrawal, or a future Aleynikov, would not need that workaround.

The episode is a case study in the interplay between courts and Congress. The judiciary read the statute as written, the reading revealed a loophole, and the legislature closed it. *Agrawal* sits at the hinge, applying the pre-amendment law one last time while the corrected statute was already on the books for conduct going forward.

## Open questions

Even after the 2012 amendment, *Agrawal* leaves questions about the reach of the general property statutes. The NSPA analysis still turns on tangibility, so the treatment of theft that produces no physical artifact but also no clean electronic transfer, for example memorization or cloud replication, remains unsettled under that statute. The decision does not resolve how the securities-in-commerce theory would apply to internal software unrelated to any traded instrument, a scenario the EEA amendment now largely addresses but the NSPA does not. And the partial dissent's disagreement over the EEA element signals that reasonable judges differed on how attenuated the connection to interstate commerce could be, a question the amendment mooted for the EEA but that can recur wherever an older or differently worded statute applies.

## Implications for inventors and businesses

- **Digital theft is now squarely covered, but litigation history matters.** The 2012 amendment closed the "produced for commerce" gap for the EEA, yet older conduct and other statutes may still turn on the medium. Understand which version of the law governs the conduct at issue.
- **Preserve the tangible trail.** *Agrawal* was convicted in part because he created physical evidence by printing the code. Access logs, print records, and device forensics that document how information left the company are often decisive in criminal referrals.
- **Internal-use software is protectable, and worth protecting.** The trading systems at issue were never sold, yet they were among the companies' most valuable assets. Treat proprietary internal tools with the same secrecy controls as saleable products, because the criminal law now reaches them.
- **Coordinate civil and criminal strategy.** A theft of code can support parallel civil DTSA claims and a criminal referral. The evidence that proves tangible or electronic transfer serves both tracks, so preserve it comprehensively from the outset.

## Frequently asked questions

**How did Agrawal differ from Aleynikov if both stole trading code?** The medium of the theft. Sergey Aleynikov uploaded Goldman Sachs source code to a server and downloaded it, an electronic transfer the Second Circuit held was not the theft of a tangible good under the National Stolen Property Act and not covered by the EEA as then written. Samarth Agrawal printed thousands of pages of Societe Generale's code and carried the paper home. The Second Circuit held the paper was a tangible good satisfying the NSPA, and that the EEA element was met because the code related to securities traded in interstate commerce.

**Did Agrawal overrule Aleynikov?** No. The panel distinguished *Aleynikov* rather than overruling it. *Aleynikov* remained good law on its facts, that purely electronic transfer of intangible code did not satisfy the statutes as then written. *Agrawal* turned on the different fact that the defendant reduced the code to physical paper, which the court treated as the theft of a tangible thing. The two decisions together mapped a line drawn by the medium of the theft.

**What did Congress do in response to these cases?** After *Aleynikov* exposed the gap, Congress passed the Theft of Trade Secrets Clarification Act of 2012, amending 18 U.S.C. 1832 to cover trade secrets related to a product or service used in or intended for use in interstate or foreign commerce, not just products produced for such commerce. That change was meant to ensure internal software like trading systems is covered regardless of the medium of theft, closing the loophole the two prosecutions revealed.

## Authorities and sources

- *United States v. Agrawal*, 726 F.3d 235 (2d Cir. 2013), slip opinion (No. 11-1074-cr): [govinfo](https://www.govinfo.gov/app/details/USCOURTS-ca2-11-01074)
- *United States v. Agrawal*, opinion (PDF): [Crowell & Moring](https://www.crowell.com/a/web/uAsuPoQLWq7ZJS8i2XwJf1/4Ttiwh/United-States-v-Agrawal.pdf)
- Crowell & Moring, "Second Circuit Affirms Conviction for Theft of Computer Code": [client alert](https://www.crowell.com/en/insights/client-alerts/second-circuit-affirms-conviction-for-theft-of-computer-code-under-federal-criminal-trade-secrets-statutes)
- *United States v. Agrawal*, case overview: [Wikipedia](https://en.wikipedia.org/wiki/United_States_v._Agrawal)
- 18 U.S.C. 1832 (EEA, theft of trade secrets, as amended 2012): [Cornell LII](https://www.law.cornell.edu/uscode/text/18/1832)
- 18 U.S.C. 2314 (National Stolen Property Act): [Cornell LII](https://www.law.cornell.edu/uscode/text/18/2314)