Prediction markets have never been busier. According to Bloomberg News, weekly trading volume across platforms such as Polymarket and Kalshi surpassed $2 billion in late October 2025—an all-time high that even eclipsed the record levels seen during the 2024 presidential elections. A significant portion of that surge came from bets tied to the New York City mayoral race, where the names Andrew Cuomo and Zohran Mamdani dominated both digital order books and political conversation.
The extraordinary activity has drawn global attention. Not only because the race itself has become a national proxy for broader debates over class, education, and identity, but also because the markets meant to measure public sentiment increasingly seem to shape it.
Prediction markets were originally presented as rational forecasting tools, systems designed to crowdsource dispersed information and translate it into measurable probabilities. That academic optimism remains partly justified: such platforms can indeed highlight collective insight and even outperform traditional polling. Yet under the same banner of innovation, another pattern has emerged. A mechanism built to quantify uncertainty can, in certain contexts, also exploit it.
From Insight To Exposure
When used responsibly, prediction markets can inform decision-making across science, policy, and economics. But as access widened, they migrated from academic and institutional settings into retail-trading apps and social networks. In late October, Truth Social announced ‘Truth Predict,’ a partnership with Crypto.com to embed real-money forecasting directly into its social platform—a move that further blurs the boundary between political speech and speculative play.
These markets are designed for engagement. Fast feedback, variable rewards, and continuous updates sustain user attention, just as they do in gaming or social media. Yet those same design features can create cycles of compulsion.
Understanding The Risk
Most people underestimate what makes gambling systems harmful. It is not only the possibility of loss but the psychology of recovery. Each failed trade invites an effort to “win it back,” pulling participants into rapid, reactive, and riskier transactions aimed at minimizing short-term losses. Over time, this pattern produces impulsivity, emotional volatility, and, in some cases, serious financial harm. Online prediction platforms deliberately engineer these vulnerabilities through manipulative design features: auto-play functions, ambient notifications, countdown timers that manufacture urgency, and deceptive interface techniques that make protective measures like deposit limits difficult to access. These systems exploit psychological biases through immediate social comparison feedback and variable reward schedules, tapping into drives for achievement and social belonging while creating continuous engagement loops that can outpace deliberate decision-making. Such mechanics, while familiar in casinos, now appear in financial interfaces marketed as information tools. The crossover between trading, prediction, and social engagement introduces behavioral risks that regulators have barely begun to address. The boundaries are not just blurring—they’re being deliberately erased. DraftKings’ pricey acquisition of Railbird to enter prediction markets, alongside the New York Stock Exchange’s reported stake in Polymarket, signals a new era where sports betting platforms processing hundreds of millions in weekly volumes can pivot seamlessly into political and financial prediction markets. In this new era, the convergence of financial regulation, gambling oversight, and public health protection has become urgent. In our academic work, we have examined this intersection, where financial regulation, gambling oversight, and public health converge, and traditional boundaries grow increasingly porous. But beyond personal harms, these markets can create broader informational vulnerabilities that affect the collective sphere.
From Private Impulse To Public And Elections’ Influence
What begins as an individual feedback loop can easily scale into a collective one. The same emotional dynamics that drive users to over-trade can also amplify narratives in the political realm, where attention itself functions as capital.
In the past month, online discussions of the New York City mayoral race illustrated this crossover vividly. Investor Bill Ackman and commentator Mark Moran raised questions amid online discussions about whether certain trading patterns might be influencing perceptions of political momentum. Screenshots of fluctuating odds on Klashi and Polymarket circulated widely across X (formerly Twitter), where audiences and journalists alike treated them as evidence of real-world shifts.
Reports also raised questions about whether parts of the unusual volume reflected coordinated timing by well-resourced participants, including possibly non-U.S. investors, or entities with policy interests. While such claims remain unverified, they underscore a growing vulnerability: when prediction markets intersect with politics, the distinction between data aggregation and influence operation becomes dangerously porous. Even the perception that external actors might be shaping sentiment through financial signals can erode public trust in both markets and democratic discourse.
A self-reinforcing cycle followed: conversation drew attention; attention moved market prices; and updated prices circulated as confirmation. This dynamic—where perception reshapes reality—is what researchers call reflexivity. The market ceases to mirror opinion and begins to manufacture it.
Designing Conviction
Platforms like Polymarket and Kalshi are not neutral venues for information exchange. Their interfaces borrow directly from gaming and gambling: bright colors, animated feedback, and “win” effects that deliver momentary satisfaction independent of accuracy. Each trade provides a sense of reward, reinforcing confidence rather than reflection. For some users, this enhances engagement. For others, it blurs the line between insight and entertainment.
Thin Markets, Strong Signals
Unlike large financial exchanges, event-based platforms often have limited liquidity. A small number of high-value traders can significantly move displayed probabilities, and algorithms designed to spotlight “trending” activity amplify those movements further. The result is a structure highly susceptible to coordinated or concentrated influence—even if not illegal, such dynamics can make collective data appear more certain or democratic than it really is. While most are betting on outcomes, a few shape the narrative, creating the illusion of inevitability.
Regulatory Gray Zones
Oversight of prediction markets remains fragmented. Federally, the Commodity Futures Trading Commission (CFTC) oversees certain event-based contracts, while state gaming and consumer-protection agencies claim overlapping authority. The outcome is a patchwork that allows risk to migrate between jurisdictions.
The platform Kalshi illustrates this tension. In 2024, a federal district court in Washington, D.C. ruled that Kalshi could proceed with political-event contracts after the CFTC denied its application. The agency then appealed to the D.C. Circuit Court of Appeals, but later withdrew the case in May 2025, leaving Kalshi able to operate without a definitive ruling or long-term framework. That regulatory uncertainty has not dampened enthusiasm: in October 2025, Kalshi raised $300 million at a $5 billion valuation, signaling investor confidence in a market that remains legally unsettled.
Kalshi now describes itself as CFTC-compliant for specific product categories while continuing to test the boundaries of permissible activity. Its position exemplifies a broader trend in fintech—companies expanding in regulatory gray zones where innovation outpaces enforcement.
Probabilistic Perceptions, Elections, And Democratic Discourse
The democratic implications are profound. Market odds, especially when amplified on social media, shape perceptions of who is leading or lagging. Those perceptions influence donations, media framing, and turnout expectations. A colorful chart showing “probabilities” can influence public attention in ways comparable to campaign messaging.
No conspiracy is required for this effect to matter. Once the infrastructure exists, it rewards the most attention-grabbing narratives, amplifying bias and emotion under the guise of data.
Managing Harm, Preserving Value
Mitigating these risks requires coordination, not prohibition. Platforms should disclose large positions, the timing of major trades, and any affiliations relevant to public events. Regulators can extend existing consumer-protection tools—spending caps, self-exclusion options, and “cool-off” periods—to event-based platforms accessible to the general public. Viral market data should be labeled as indicative, not factual.
Ultimately, education is essential. A figure claiming a “78 percent chance of victory” is not a statement of fact but a snapshot of sentiment and liquidity. Understanding that difference is vital to maintaining both market integrity and democratic trust.
Prediction markets retain potential as instruments of collective intelligence. But without transparency and behavioral safeguards, these markets risk turning foresight into persuasion and data into propaganda.
This prediction markets & elections Op-Ed was co-authored with Dr. Sharon Rabinovitz, Head of the Graduate Program in Crime & Addiction Studies at the School of Criminology, and Director of the Unit for Excellence in Research & Study of Addiction (ERSA) at the Faculty of Law, University of Haifa.