Andreessen Horowitz (a16z) is making a bold bet that prediction markets, combined with artificial intelligence and blockchain technology, could become essential infrastructure for the digital economy. The venture capital firm submitted a formal comment to the Commodity Futures Trading Commission (CFTC) urging regulators to develop a regulatory framework that allows prediction markets to scale beyond their current niche status.

What Are Prediction Markets and Why Does a16z Think They Matter?

Prediction markets are platforms where people trade contracts based on the likelihood of future events occurring. Think of them as betting markets, but with a serious economic purpose: they aggregate dispersed information from many participants into prices that reflect collective knowledge about what is likely to happen. From September to March of this year, trading volume on Kalshi, one of the largest prediction market platforms, surged tenfold, from $300 million to $3 billion in average weekly volume.

a16z argues that prediction markets could solve real business problems that currently lack good solutions. A manufacturer exposed to commodity price risk tied to geopolitical events, a retailer whose revenue depends on weather patterns, or an insurer seeking more granular hedging tools could all benefit from direct prediction markets rather than relying on crude proxy assets.

How Could AI and Blockchain Transform Prediction Markets?

a16z's vision goes beyond current prediction market applications. The firm outlined a convergence of three technologies: prediction markets, artificial intelligence, and blockchain infrastructure. AI agents are already capable of processing vast amounts of information and executing trades autonomously. Blockchain networks offer permissionless, programmable infrastructure that these agents can interact with directly. Together, they could enable a future where AI agents autonomously manage business risk across event contracts offered on blockchain-based prediction markets, continuously adjusting positions as conditions change without human intermediation.

In this scenario, prediction markets become more than just risk-management tools. They transform into what a16z calls "information machines," mechanisms for aggregating dispersed information into prices that reflect collective knowledge. When AI agents participate in these markets at scale on open blockchain infrastructure, the result is a powerful system for determining what is likely to be true.

What Regulatory Roadblocks Are Preventing Market Growth?

a16z identified several policy barriers that currently limit prediction markets from reaching their potential:

• State-Level Conflicts: A patchwork of conflicting state laws and regulations, along with uncoordinated enforcement actions and cease-and-desist letters from state attorneys general, threatens the viability of prediction markets and prevents residents from accessing them.
• Contract Resolution Issues: Prediction markets deal in uncertainty, and disputes sometimes arise about whether an event underlying a contract actually occurred, limiting the markets' ability to scale without better resolution mechanisms.
• Manipulation and Inside Information: When individuals who can directly influence event outcomes breach duties of trust, overall trust in prediction markets decreases, requiring more effective market surveillance.
• Ineffective Special Rules: The CFTC's 2011 rulemaking imposed a blanket ban on event contracts in certain categories without conducting required public interest determinations, preventing legitimate market development.
• Blockchain Clarity Gaps: Current regulatory principles create uncertainty about whether and how blockchain technology can be used to lower transaction costs, improve access, and enable 24/7 availability.

a16z emphasized that Congress created a uniform national regulatory framework for Designated Contract Markets (DCMs) to avoid exactly the kind of state-level fragmentation currently plaguing prediction markets. The firm argues prediction markets need the same consistent treatment.

Steps for Regulators to Support Prediction Market Infrastructure

a16z outlined specific recommendations for how the CFTC should approach prediction market regulation:

• Reaffirm CFTC Jurisdiction: Congress clearly intended to preempt state laws that aim to license or prohibit a DCM's operation within state borders, and binary option event contracts are swaps subject to exclusive CFTC jurisdiction.
• Conduct Public Interest Determinations: Rather than imposing blanket bans on event contract categories, the CFTC should conduct the public interest analysis required by the Commodity Exchange Act for each category.
• Develop Blockchain-Friendly Rules: Regulators should clarify how blockchain technology can be used in prediction markets to unlock benefits like improved access, lower costs, and continuous availability.
• Establish Market Surveillance Standards: Create effective coordination and market surveillance mechanisms to root out manipulation and insider trading while protecting legitimate market participants.

The CFTC issued an Advance Notice of Proposed Rulemaking (ANPRM) on prediction markets earlier this year as part of Chairman Selig's effort to promote responsible innovation in derivatives markets. a16z applauded this step, noting that the agency has overseen these markets for more than two decades but this ANPRM signals a reinvigorated commitment to getting the regulatory framework right.

a16z stressed that the rules developed now will determine whether prediction markets can evolve into broadly useful infrastructure for risk management and information discovery. "The technology should not be constrained in its infancy by rules designed without regard for where it is heading," the firm stated in its comment letter.

The stakes are significant. In a world where misinformation is pervasive and trust in institutions is declining, markets that synthesize human and machine intelligence into reliable probability signals could become critical information infrastructure for the internet itself. But none of this happens if the regulatory framework treats prediction markets as a niche product to be narrowly tolerated rather than as foundational infrastructure worth supporting at scale.