Prediction markets have long promised to aggregate information about future events. Increasingly, these signals come not only from people, but also from machines.
Autonomous AI agents are emerging as powerful tools for trading prediction markets, especially for retail users trying to compete in an increasingly automated environment, according to David Minarsch, CEO and co-founder of Valory AG, the team behind crypto-AI protocol Olas.
Valory builds products at the intersection of blockchain and multi-agent systems (MAS), and is currently focused on Olas, formerly known as Autonolas. The protocol is designed as an infrastructure for autonomous software agents that can run services on blockchains, interact with smart contracts, and cooperate with each other while earning cryptographic rewards.
The broader vision is what Minarsch calls an “agent economy.” A decentralized ecosystem where autonomous AI agents perform useful tasks and generate value for their users.
One of the most visible experiments in this vision is Polystrat, an AI agent launching on market prediction platform Polymarket in February 2026. The agent trades on behalf of its custodial and ownership users, executing strategies continuously around the clock.
“In a nutshell, Polystrat is an autonomous AI agent that trades on Polymarket 24/7 on behalf of its human user,” Minarsch said. The idea is simple: while humans sleep, work, or lose concentration, the agent continues to negotiate.
Prediction markets, platforms on which users trade contracts tied to actual outcomes, have evolved from niche forecasting tools to a growing fintech sector in recent years. The industry’s defining moment came during the 2024 US presidential election, when trading volumes soared and markets gained visibility, followed by rapid expansion in sports, economics and crypto-related betting. By 2025, the total notional trading volume on major platforms exceeded $44 billion, with monthly activity reaching up to $13 billion during peak periods.
Today, the market is heavily concentrated around two dominant players: Kalshi, an event-driven contracts exchange regulated in the United States and overseen by the Commodity Futures Trading Commission, and Polymarket, a crypto-native platform that operates globally and offers a wider range of prediction markets. Together, they account for approximately 85-97% of trading volume. in the industry, handling tens of billions of dollars in annual bets on everything from elections and central bank policy to sporting and cultural events
Why machines can outperform humans
The push toward AI-driven trading stems from a simple observation. Much of the intelligence built into modern AI models has not yet translated into financial markets.
This realization prompted the Valory team to begin building what it calls a “prediction market economy” on Olas in 2023, an ecosystem in which AI agents use prediction tools and data pipelines to predict outcomes and trade them.
Prediction markets themselves rely on probabilistic forecasts. A simple guess about an event, whether a political outcome, an economic indicator or a sporting outcome, may be no better than a coin toss. But structured data analysis and disciplined trading strategies can change this equation.
“Simply offering ready-made models with markets generally does not result in results better than a coin toss,” Minarsch said. “But cutting-edge AI models integrated into custom workflows, called prediction tools, have historically shown predictive accuracy of up to 70% and above.”
The results so far suggest that the machines may have an advantage. Third-party data indicates that only about 7-13% of human traders achieve positive performance in prediction markets, while the majority lose money.
At the same time, machine participation is increasing rapidly. More than 30% of wallets on Polymarket already use AI agents, according to analytics platform LayerHub.
Minarsch believes this trend reflects a broader shift: Humans are already competing with machines, whether they realize it or not. “There are human participants in prediction markets alongside many machines,” he said. “So humans are already engaged in a battle against machines. »
The main difference is that machines are less emotional and better able to stick to consistent strategies.
By making AI agents available to everyday users, Olas aims to level the playing field.
Early Traction for Self-Employed Retailers
Polystrat’s first performances have been encouraging.
About a month after its launch, the agent executed over 4,200 trades on Polymarket and saw unique returns of up to 376%, according to data shared by the team.
“Agents tend to do better than humans,” he said. “Polystrat AI agents are already outperforming human participants in Polymarket, with over 37% of them showing a positive P&L, compared to less than half that figure for human participants.
Users can configure their own agents based on their policy preferences, data sources or risk tolerance.
The long tail of predictive markets
Beyond performance, Minarsch believes that AI agents could unlock a neglected opportunity in prediction markets: the “long tail” of niche or localized questions.
Many prediction markets revolve around major world events, elections, macroeconomic data or high-profile sporting competitions. But countless small questions remain largely unexplored.
“Humans often don’t bother to look for information,” Minarsch said. “They don’t bother to make the effort.” In contrast, AI agents can analyze a large number of small markets simultaneously.
“The long tail of prediction markets is very interesting for AI agents,” he said. “Just point the agent at the problem and he does the job. »
This could help expand prediction markets as a data collection tool for businesses, policy makers. Forecasting markets have long been studied as a way to bring together scattered knowledge and surface insights that traditional surveys or models might miss.
In this sense, prediction markets can become a kind of upstream technology for decision-making across industries.
Human-AI collaboration
Despite the rise of automation, Minarsch doesn’t see AI agents replacing humans entirely.
Instead, he presents them as complements.
“Humans make more hasty choices, which can be detrimental,” he said. “AI agents can act like something humans rely on.”
A future direction is to allow users to enrich their agents with proprietary knowledge or specialized datasets. “We are seeing demand from users who want their agent to leverage their own knowledge base or proprietary information,” Minarsch said. “This would allow agents to trade in a more principled way than a human being.”
Over time, the team says the prediction models and data pipelines that power these agents have improved significantly, generating sustained alpha when combined with large general-purpose language models.
Risks and regulations
The growth of prediction markets also raises ethical and regulatory questions.
Some critics argue that markets predicting war, death, or disaster could provide incentives to manipulate outcomes or profit from bad events.
Minarsch acknowledged that careful safeguards are necessary.
“There needs to be regulation on the types of prediction markets that should exist,” he said.
At the same time, he believes AI agents could also help detect problematic markets or manipulation attempts by identifying suspicious patterns.
“Agents could spot trends and help close problem markets,” he said.
Building a user-owned AI economy
For Minarsch, the ultimate goal is not simply better trading strategies.
This is about ensuring that everyday users retain their share in an increasingly automated digital economy.
A future in which AI systems carry out the bulk of economic activity could risk disenfranchising individuals if centralized platforms control the technology.. “Olas aims to create a world in which human users can be empowered through their AI agents rather than disenfranchised. »
To counter this dynamic, the project emphasizes user ownership of AI systems. “We want to create more user-owned agents,” Minarsch said.
If successful, this model could enable users to deploy autonomous software that generates value on their behalf across markets and services. Prediction markets are just the starting point.
Learn more: AI rout hits software stocks, but Grayscale says blockchains will benefit
