What Happened
OpenAI fired an employee in late February 2026 for making trades on prediction market platforms, specifically Polymarket and Kalshi, using non-public information about the company. Prediction markets allow users to bet on the outcomes of future events. In this case, the employee reportedly traded on events related to OpenAI product releases or company decisions where they had advance knowledge. Wired confirmed the firing and its connection to prediction market trading behavior.
Prediction markets have grown significantly in mainstream visibility following the 2024 US election cycle, which drove substantial trading volume and brought both platforms to wider public awareness.
Why It Matters
Prediction markets occupy a legal grey area. They are not regulated as securities markets in the US, which means traditional insider trading prohibitions do not automatically extend to them in the same way they apply to stock markets. OpenAI's decision to fire the employee despite this legal ambiguity signals that the company treats the behavior as an ethics violation regardless of strict legal exposure.
As prediction markets grow in volume and legitimacy - Polymarket and Kalshi now process tens of millions of dollars in trading on major events - the question of whether tech employees can trade on inside knowledge about their own companies becomes more material to the industry. OpenAI, Anthropic, Google DeepMind, and other AI labs where product announcements visibly move prediction market prices are all potentially affected by how this norm gets established.
The precedent also matters for smaller AI companies whose funding rounds, model releases, and partnership announcements are actively traded on these platforms.
Our Take
OpenAI handled this correctly: act on the ethics violation and let the legal questions sort themselves out later. The behavior is a clear misuse of privileged information regardless of whether existing law provides a specific remedy. Waiting for regulatory clarity before enforcing internal standards would send the wrong signal to the rest of the organization.
For employees at AI companies: prediction markets are not currently regulated as securities, but that is changing. The CFTC is actively reviewing its oversight of prediction markets, and several members of Congress have pushed for expanded regulation. Trading on inside knowledge may be technically permissible today and clearly illegal in 18 months. Treating them as a grey area to exploit is a bet on regulatory stagnation that seems increasingly unlikely to pay off. The OpenAI firing gives other AI companies a clear signal that they should establish explicit policies on prediction market trading before the next incident forces the conversation at a worse time.