Crypto prediction markets are rapidly being reshaped from a human-centred realm of judgment into an automated financial ecosystem using autonomous artificial intelligence agents.

On March 15, local time, blockchain media outlet CoinDesk reported that Valory, the developer of the crypto AI platform Olas, is stepping up a prediction-market trading model using AI agents and ushering in what it calls the era of the Agent Economy. Valory provides infrastructure that allows blockchain-based autonomous software agents to interact with smart contracts, execute strategies around the clock and manage users' assets on their behalf.

At the centre of this shift is Polystrat, an AI agent launched on prediction market platform Polymarket in February. Polystrat is an autonomous agent owned and managed directly by users. It analyses market conditions and executes trades even while humans sleep or work. Valory said Polystrat executed more than 4,200 trades within a month of launch and posted returns of 376 percent on some single trades, outperforming human participants.

Analysis platform LayerHub said more than 30 percent of wallets on Polymarket are already using AI agents. Data also showed that only 7 to 13 percent of human participants make profits, while more than 37 percent of AI-agent users recorded positive profit and loss, showing machines have gained an edge over humans in prediction markets. That is because machines can maintain consistent strategies without being swayed by emotion, a trait that acts as a strong competitive advantage in fast-changing financial markets.

David Minarsch (데이비드 미나르슈), Valory's chief executive officer, forecast that the use of AI agents will be maximised not only for major global events but also in niche markets in the long tail, which are hard for humans to track one by one. He explained that because AI agents can process vast amounts of data simultaneously that humans struggle to grasp, prediction markets could be expanded into data-collection tools that provide real-time insights to companies or policymakers.

As market automation accelerates, ethical and regulatory challenges are also coming into focus. With criticism of prediction markets that deal with sensitive topics such as war, death and disasters, Minarsch acknowledged the need for appropriate guidelines and regulations. At the same time, he stressed that AI agents can also play a positive role by detecting suspicious patterns or manipulation attempts in the market and identifying and cleansing improper trades.

In this context, Olas is continuing to develop technology so users can combine their own proprietary data or knowledge bases with agents to execute more advanced trading strategies. Valory's ultimate goal is to support AI technology that is not dependent on centralised platforms, enabling individual users to create value and maintain their rights and interests in an autonomous economy through AI agents they own directly.

This prediction-market automation experiment is expected to become a first test bed, going beyond a simple trading profit model, as humans and autonomous machines coexist and adapt to a new digital economic environment.