X user 0x_Discover said the AI alerted him at 3:47 AM to deploy $12,000 (about Rs 11.08 lakh) across six markets.
The trades reportedly returned $43,800 (around Rs 40.47 lakh) within hours after the markets resolved.
The AI monitored global data sources such as government feeds, parliamentary livestreams and central bank updates to detect pricing gaps.
While most of the users around the world are using AI to write code, articles, plan trips and research, an X user has used it to double down on the corpus within a few hours. An X user has claimed that an AI agent has helped him gain overnight profits by spotting time-zone arbitrage opportunities on the prediction platform Polymarket.
Taking to X, a user named 0x_Discover claimed that an AI agent alerted him in the early hours of the morning about a set of markets that were close to resolution. According to the post, the notification arrived at around 3:47 AM, requesting approval to deploy $12,000 (about Rs 11.08 lakh) across six prediction markets expected to settle within the next 90 minutes.
The user said he approved the trade and went back to sleep. By the time the markets resolved later in the morning, the system had reportedly generated profits of $43,800, which is roughly Rs 40.47 lakh.
Talking about how it happened, the user stated that the trading agent was built using an AI framework called OpenClaw and had been running for about nine days. Its primary purpose was to identify time zone arbitrage opportunities on Polymarket, situations where the real-world outcome of an event becomes known in one region while market prices have not fully reacted elsewhere.
To get this, the agent was connected to several global data sources. These sources include – Japanese government RSS feeds, livestreams of the European Parliament, Australian financial news wires, Middle East flight tracking data and announcements from Asian central banks.
The system then followed a simple rule- sending an alert whenever a market was scheduled to resolve during US night hours, specifically between 2 AM and 6 AM Eastern Time. As per the post, the AI flagged six such markets that night. He also shared the examples, which included signals around a Japanese interest rate decision, a vote in the European Parliament, a policy update in South Korea, an Australian trade announcement, production-related notes from OPEC involving the United Arab Emirates, and a regulatory vote in Singapore.
The user also stated that the trades were entered at prices between 15 cents and 31 cents, while the final payouts ranged between 95 cents and $1 once the markets resolved. The post argued that the strategy works because many traders on Polymarket are based in the United States.
Of course, the claim has sparked a massive buzz around on X with users talking about whether AI agents can actually help in trading strategies and predictions.
