rewrite this title Prediction markets hit $64 billion in 2025 but reliance on centralized logins has created a critical security flaw

Prediction markets entered the mainstream in 2025, with a fourfold surge in annual trading volume as a handful of venues consolidated control over what is rapidly becoming an institutional-scale product, according to a new report from blockchain security firm CertiK.

The sector’s total volume rose from $15.8 billion in 2024 to $63.5 billion in 2025, the report said, with activity remaining elevated after the US election cycle and extending into January 2026.

That persistence matters because it suggests that election trading behaved less like a one-off spike and more like an acquisition event that drew new users into repeat behavior.

Notably, the week ending Jan. 18 set a record of about $6 billion in notional volume, the report said, reflecting how quickly prediction markets have moved from niche crypto product to a high-turnover trading venue.

However, CertiK’s central argument is that the next phase of growth is colliding with an integrity problem that has less to do with smart contract exploits than with the layers that govern onboarding, the “real” meaning of volume, and the mechanisms that determine who gets paid.

A three-platform market with single-point failures

Three platforms now account for more than 95% of global prediction market volume, according to CertiK, and each is pursuing a different path to dominance.

Kalshi, which operates as a regulated venue in the US, is positioned as the compliance-first model. Polymarket has captured the largest share of crypto-native and international participation.

Meanwhile, Opinion is the fast-growing entrant, using ecosystem incentives to scale from effectively zero to roughly 30% market share in months, the report said.

That concentration turns operational issues into systemic ones.

A failure at any major venue is no longer a contained event; it is a market-wide trust shock that can spread across liquidity pools, data feeds, and user balances, particularly as brokers and mainstream distribution begin to treat prediction probabilities like a new class of information product.

CertiK points to a December 2025 incident involving Magic.link, Polymarket’s third-party authentication provider, as a preview of where the sector is most exposed.

Accounts using Web2-style login methods, such as email or social authentication, were compromised, placing funds in affected accounts at risk, while the on-chain settlement layer remained secure.

In CertiK’s framing, it was an identity failure, not a settlement failure, and it highlighted the tradeoff of “Web2.5” onboarding: a smoother user experience in exchange for centralized failure points.

The lesson is uncomfortable for an industry that markets itself on decentralization.

Prediction markets can support fully collateralized on-chain settlement while retaining the same third-party risks that plague conventional fintech, including authentication, account recovery, and platform-level access controls.

When the tape lies but the odds still talk

The report also draws a line between two concepts that are often conflated in crypto markets: trading volume as a proxy for adoption and probability outputs as a proxy for information.

According to the report, incentive programs can inflate activity without necessarily improving the quality of forecasting signals.

CertiK reported that wash trading remains widespread, citing research estimating that artificial volume reached as high as 60% on some platforms during peak airdrop-farming periods.

Such distortion can mislead outsiders, including prospective institutional users, regarding liquidity depth and organic participation.

Yet CertiK argues the more important question is whether the probabilities remain useful even when the tape is noisy.

In the report’s view, wash trading has inflated volume metrics but has not yet compromised price accuracy, and probability outputs have remained reliable for forecasting.

This creates tension for platforms seeking to graduate to mainstream finance; they may be able to position themselves as information utilities even if their activity metrics are partly fabricated by incentives.

It also raises a harder strategic decision for the market leaders.

If distribution and credibility depend on information quality, platforms may have to become less tolerant of behaviors that boost volume in the short term but undermine the optics and trust required for institutional capital.

Chain migration and the new execution plumbing

Beneath the headline numbers, CertiK describes a structural rotation in how prediction market liquidity is executed.

Polygon retained “legacy dominance” through the November election cycle, the report said, but BNB Chain volume surged beginning in late 2025, correlating with Opinion’s accelerated incentive rollout.

By the week of Jan. 19, CertiK said BNB Chain activity had effectively flipped the historical hierarchy, capturing the plurality of weekly flows and pushing off-chain settlement into a secondary position, even as Kalshi posted record performance during NFL playoff trading.

That shift is more than a scoreboard for blockchain ecosystems. It changes who can participate, how trades are cleared, and which market structures are feasible.

CertiK notes that many on-chain venues are moving from automated market makers to central limit order books deployed directly on high-throughput chains, a design that produces tighter spreads and more familiar mechanics for professional traders.

In practice, it also moves prediction markets closer to an exchange-like microstructure, with the attendant risks of front-running and the MEV-style transaction-ordering disadvantages on public networks.

The oracle problem, the moment where “truth” becomes a payout

If there is a single tail risk that unifies the sector’s growth story, it is resolution, the step that converts probabilities into cash.

CertiK characterizes oracle manipulation as the primary technical attack vector because market-resolution mechanisms directly control fund distribution.

It also says ambiguous market definitions have already caused disputes across all major platforms throughout 2025, especially where political outcomes or contested official results create gray areas.

The report maps the main resolution models across the dominant platforms.

Polymarket is described as using UMA’s optimistic oracle, in which outcomes resolve automatically unless disputed within a challenge window, with disputes escalating to UMA token-holder votes.

Kalshi is framed as using centralized arbitration, with human arbiters resolving outcomes based on authoritative sources.

Opinion is described as relying on consensus oracles, where designated parties must agree on an outcome.

Each model carries a different trust assumption. Optimistic oracles can be fast for unambiguous outcomes but create edge-case vulnerability, including the risk that large token holders may influence votes in low-liquidity disputes.

Centralized arbitration is predictable but requires trusting the platform operator. Consensus oracles distribute authority but still depend on the incentives and integrity of the designated resolvers.

As prediction markets scale, those tradeoffs become harder to ignore.

The sector can tolerate occasional edge-case controversy when it is a crypto curiosity. However, becomes a governance crisis when market probabilities begin to appear in mainstream distribution channels or are used by institutions as inputs to risk decisions.