Should confidential tech-company data used to profit on prediction markets be regulated as insider trading, market manipulation or something else?

Advocate A’s weakest move is to say the “policy principle is the same” as securities insider trading and then treat the label as secondary. That is exactly where the legal and conceptual risk lies. The facts, if proved, look like a serious breach of Google confidentiality and potentially a fraud; but the legal theory cannot simply be imported from stock-market insider trading because the traded instrument was not Google stock, not a claim on Google cash flows, and not a security whose price is supposed to reflect issuer financial information. The DOJ complaint itself confirms this: Count One is commodities fraud under the Commodity Exchange Act and CFTC Rule 180.1, Count Two is wire fraud, and Count Three is money laundering. There is no securities-fraud count. That is not a technicality; it shows prosecutors know they need a commodities/swap hook or a property-fraud hook, not generic “insider trading.”

A’s CFTC point is partly right but incomplete. The CFTC’s 2022 Polymarket order treated Polymarket’s event-based binary options as swaps when offered off-exchange without DCM or SEF registration. If the Google Year in Search contracts are validly swaps or commodity options under the CEA, then CFTC antifraud authority may apply. But that supports a bounded commodities-fraud argument, not an open-ended rule that confidential data used in any prediction-market bet is “insider trading.” Prediction markets cover elections, sports, pop culture, corporate announcements, wars, weather, and entertainment. Many are designed to reward superior information. The line must be breach plus a covered instrument plus a statutory antifraud theory, not merely “nonpublic advantage.”

The second missing constraint is property. The Second Circuit’s 2025 Chastain decision, involving OpenSea NFT “insider trading,” vacated a wire-fraud conviction and held that confidential business information must have commercial value as property; unethical trading on internal information was not enough if the jury could convict on ethics alone. That precedent matters here because it pushes courts away from headline-driven “insider trading” analogies and toward proof that the misused information was Google’s commercially valuable property and that the statute charged covers the trading conduct. The DOJ complaint tries to meet that standard by alleging Google’s Year in Search data was commercially valuable, confidential, access-restricted, and central to a marketing campaign. That may be enough for wire fraud, but it is not the same as material nonpublic information about an issuer’s securities.

The strongest proper regulatory path is therefore narrower than A’s: prosecute theft or misappropriation through wire fraud if the data is property, through commodities fraud only if the contracts fall within CEA jurisdiction, and through platform or CFTC event-contract rules for market integrity. Calling every such case “insider trading” risks criminalizing legitimate informational advantages in polling, web analytics, journalism, sports modeling, and alternative data unless Congress or the CFTC defines prediction-market-specific duties ex ante.

A’s rebuttal concedes the decisive point: this case cannot rest on a free-floating norm that “confidential data plus prediction-market profit equals insider trading.” A says enforcement must be tied to “breach of a duty, commercially valuable proprietary information, a covered trading instrument under the CEA theory, and deception for personal gain.” That is not a refinement of A’s position; it is B’s position. The concession implies that “insider trading” is at most a headline shorthand. The legally operative questions are whether the Polymarket contracts are swaps or otherwise covered CEA instruments, whether Google’s information is property under wire-fraud doctrine, and whether the defendant engaged in deception or misappropriation within those statutes.

A’s remaining evidence proves alleged wrongdoing, not the broader doctrinal conclusion. The criminal complaint charges commodities fraud under 7 U.S.C. § 9(1), CFTC Rule 180.1, wire fraud, and money laundering; it does not charge securities fraud. The DOJ press release likewise describes the conduct as misappropriation of confidential business information used for Google-related trades on a prediction-market platform, while listing Commodity Exchange Act, wire-fraud, and money-laundering counts. Those facts support prosecution if proved. They do not support importing securities insider-trading concepts into prediction markets generally.

A also overstates the CFTC point. The 2022 Polymarket order found that Polymarket’s event markets, composed of paired binary options, constituted swaps under CFTC jurisdiction when offered off-exchange. That gives the CFTC a possible statutory hook for event contracts fitting that structure. It does not mean all prediction-market bets are conceptually equivalent to securities trades, or that any confidential corporate, polling, journalistic, web-analytics, or sports data used to forecast an event should be governed by insider-trading norms. The correct limiting principle is not A’s “insider-style market abuse”; it is the statutory boundary: covered instrument plus fraud or misappropriation.

Chastain reinforces why that boundary matters. The Second Circuit held that not all confidential information qualifies as property for wire fraud; confidential business information must have commercial value, and courts must look to whether the business considered and treated it as confidential. The Google complaint tries to plead those elements by alleging commercial value, access restrictions, and confidentiality banners. But that pleading move confirms B’s argument: the case should turn on property fraud and CEA coverage, not on an elastic insider-trading label.

A’s formulation also does not adequately address the core policy risk. Prediction markets are supposed to aggregate lawful informational advantages. If regulators use “insider trading” as the master concept whenever nonpublic data predicts an outcome, the limiting line becomes opaque. Bespoke event-contract rules, CFTC antifraud authority for covered swaps, computer-access controls, confidentiality enforcement, and wire-fraud doctrine are sufficient tools without expanding securities-style insider trading beyond its legal and conceptual domain.

A’s strongest argument is that this is not ordinary prediction-market research but alleged misappropriation by an employee of the very institution whose confidential announcement determined the contract outcomes. If Spagnuolo used Google-confidential Year in Search data to trade at least twenty-three Polymarket contracts with near-perfect accuracy and $1.2 million in profit, A is right that the conduct threatens market integrity and should not be dismissed as merely clever forecasting. A also fairly limits the theory to duty, misappropriation, commercial value, deception, and a covered instrument rather than claiming that every private informational edge is illegal.

My position remains stronger because once A accepts those limiting elements, the decisive category is no longer “insider trading” in the securities-law sense but statutory fraud plus property and jurisdiction. The DOJ press release says the criminal complaint charges commodities fraud, wire fraud, and money laundering, not securities fraud. That charging choice is not incidental: it reflects that Polymarket contracts are not Google shares, do not represent claims on Google cash flows, and are not priced by reference to issuer financial disclosure. The alleged wrong is not trading Google securities on Google financial information; it is allegedly using employer property or confidential business information to trade event contracts.

The CFTC’s position also supports a narrower conclusion than A’s rhetoric suggests. The CFTC alleges “insider trading” in event contracts and says Spagnuolo owed Google a duty of trust and confidence, but its jurisdiction depends on the Commodity Exchange Act and CFTC regulations. Its 2022 Polymarket order found that Polymarket’s yes/no event markets were paired binary options constituting swaps when offered off-exchange. That may be enough for CEA antifraud liability here, but it is a statutory hook, not proof that prediction markets should inherit securities insider-trading doctrine as a master concept.

Chastain shows why this distinction matters. The Second Circuit vacated the OpenSea NFT wire-fraud conviction because the jury could have convicted on unethical conduct rather than deprivation of property, and it held that confidential business information must have commercial value to qualify as property under wire-fraud law. The government may be able to prove that Google’s Year in Search data had such value and was access-restricted. But that confirms B’s framework: liability should turn on Google’s property interest, confidentiality controls, deception, and CEA coverage, not on an elastic insider-trading analogy.

The policy line should be explicit. If a polling analyst, journalist, web-traffic researcher, or sports data scientist lawfully develops nonpublic insight, prediction markets are supposed to reward that advantage unless a specific rule says otherwise. If an employee steals or misuses protected corporate data, prosecute under wire fraud, computer-access law, confidentiality agreements, CFTC antifraud rules for covered contracts, or bespoke event-contract rules. Calling the whole category “insider trading” risks obscuring the legal elements that prevent overcriminalization.