Prediction markets have made it possible to trade on increasingly specific real-world outcomes. But they’ve also opened the door for insider trading and market manipulation.
On Aug. 28, federal regulators offered a fresh example of what can happen when someone knows one of those outcomes before everyone else.
After finding that White House teleprompter operator Gabriel Perez used advance access to speeches to trade contracts tied to words or phrases the president might say, the Commodity Futures Trading Commission ordered him to disgorge $107,539 in profits and pay a $65,000 civil penalty. The agency said Perez misappropriated material nonpublic information in breach of a duty of trust and confidence. He also accepted a three-year trading ban.
Although the case had nothing to do with drug development, it exposed an information-governance problem that pharma and biotech companies need to take seriously now that prediction markets have come to the industry. And that process starts with recognizing the risk and adjusting accordingly, according to Mark Scallon, a principal at Baker Tilly who advises life sciences companies on compliance programs, auditing and risk.
Kalshi and public-intelligence company AppliedXL launched a biotech prediction-market pilot in July tied to late-stage clinical trial outcomes and FDA regulatory decisions. The companies built in safeguards, such as listing trial contracts only after enrollment closes, requiring employment verification for traders and retaining Kalshi's broader prohibition on trading by anyone holding material nonpublic information.
Here’s the important part for pharma: Prediction markets change what can be traded in the life sciences.
For instance, an investor buying biotech stock is taking a position on the whole company — its management, cash, pipeline, commercialization prospects and other risks. But a prediction market can isolate a single question, such as whether a trial meets its primary endpoint or whether the FDA approves a drug.
That’s precisely the distinction that concerns Scallon. Pharma and biotech companies already have confidentiality and insider-information controls, he said. Prediction markets do not necessarily reveal that those systems are broken. They do, however, make individual clinical and regulatory milestones more directly monetizable.
"It raises the stakes," Scallon said.
A new market for old secrets
The legal framework is more nuanced than treating anyone with advance information as an insider trader. When the CFTC adopted Rule 180.1 in 2011, it noted that derivatives markets have historically allowed trading based on information that, while technically obtained legally, could still carry material consequences in the market. lawfully obtained material nonpublic information. The rule does not categorically ban that conduct, but trading can violate it when the process involves deception or manipulation — including using information in breach of a pre-existing duty to keep it confidential or obtaining it through fraud.
The CFTC applied that framework directly to prediction markets in a February enforcement advisory, warning that misappropriating confidential information in breach of a duty of trust and confidence can constitute prohibited insider trading in event contracts.
Federal prosecutors are applying the same principle in corporate settings. In May, prosecutors charged a Google software engineer with commodities fraud, wire fraud and money laundering, alleging that he used confidential business information to make about $1.2 million trading insider Google information-related prediction markets on Polymarket. The criminal case remains an allegation, not a conviction.
For drugmakers, the practical question is therefore not just whether information is material and nonpublic. It’s where the information travels, who owes duties around it and whether existing policies and contracts make prediction-market trading an explicit part of the compliance conversation.
The blind spot may be outside the company
A late-stage drug program can spread sensitive information across a far wider network than a company's executive suite. Scallon pointed to CROs, investigators, study coordinators, central labs, biostatistics firms, technology vendors, manufacturing partners and consultants as examples of third parties that may touch information tied to a clinical or regulatory milestone.
Companies often have mature controls around internal information flows, but less direct visibility into how information is shared and handled across those outside relationships, Scallon noted. He singled out CROs as a particularly important area to examine.
The risk can be amplified for smaller biotechs, Scallon said. Emerging companies often operate with lean teams, combine legal and compliance responsibilities and rely more heavily on outside partners to run trials, analyze data or manufacture products. At the same time, a single phase 3 result or FDA decision can have an outsized impact on a precommercial company.
That combination — more outsourcing and greater dependence on individual milestones — can make third-party information governance especially consequential for smaller companies, Scallon said.
Existing controls, new questions
That doesn’t mean pharma companies need to rebuild compliance from scratch, Scallon said. Existing confidentiality policies, codes of conduct, vendor agreements and audit rights provide much of the foundation. The more immediate task is to ensure those systems explicitly account for prediction markets.
One place to start is training. Employees who understand that they cannot trade company stock while holding material nonpublic information may not automatically understand that the same confidential information can create legal, contractual or ethical problems when used to trade an event contract. Scallon said companies should incorporate prediction-market examples into training even before every policy has been formally rewritten.
Another area to review is the audit clause. Many life sciences companies already reserve the right to audit third parties but do not routinely exercise that right around information sharing, Scallon said. Prediction-market risk is a reason to examine whether vendors actually have the compliance programs, training and controls they say they have — and how sensitive clinical or regulatory information moves inside those organizations.
Scallon also argued for deeper diligence on relationships that historically may have looked low risk. A familiar U.S.-based CRO, for example, might pass a standard questionnaire without the level of scrutiny applied to a higher-risk third party in another jurisdiction. Companies may now want to ask more specific questions about information governance, access controls, employee training and the use of confidential information for trading.
That approach is consistent with broader corporate-compliance guidance emerging around prediction markets. Law firms advising companies this year have recommended updating insider trading, confidentiality and ethics policies to address event contracts directly, rather than assuming older securities-focused language will be understood to cover them.
Awareness before overhaul
Kalshi's biotech pilot includes guardrails around trial timing, employment verification and trading on material nonpublic information, but those controls do not govern how information moves inside every drugmaker, CRO, lab or vendor that touches a development program.
For pharma, that means the arrival of prediction markets may be less a reason to invent a new compliance architecture than a reason to stress-test the one already in place. Which milestones are now directly tradeable? Who can see the relevant information before it is public? What duties apply to those people? And how far outside the company does the information travel?
Scallon put the starting point simply: "The first step is awareness."
