The US Senate is actively implementing stringent regulations on artificial intelligence and prediction markets within the Pentagon's yearly policy framework. This new language, added to the National Defense Authorization Act for fiscal year 2027—valued at approximately $1.1 trillion—prohibits the military from employing artificial intelligence in nuclear weapons decision-making, limits AI-driven surveillance of US citizens, and imposes strict rules against military personnel engaging in prediction markets using classified information.
The stipulations regarding prediction markets are particularly significant for the industry. The legislation mandates that the Secretary of Defense must formulate regulations and enforcement measures to deter both military personnel and civilians from placing wagers on platforms like Polymarket or Kalshi based on privileged intelligence.
What are the key provisions of the bill? The artificial intelligence provisions stem from an initiative led by Senator Kirsten Gillibrand in June 2026. This framework lays out three critical restrictions on the Department of Defense’s use of artificial intelligence. Firstly, AI is expressly banned from influencing nuclear targeting or launch decisions; human oversight is mandatory. Secondly, AI-based surveillance tools cannot be deployed against US citizens. Lastly, there must be significant human oversight within autonomous weapon systems.
Additionally, Senators Adam Schiff, Elissa Slotkin, and Mark Kelly have supported similar initiatives advocating for stricter regulations on autonomous weapons systems. The restrictions on prediction markets specifically address rising concerns surrounding insider trading practices within these markets. Reports indicate military personnel have engaged in speculative betting linked to major geopolitical developments. This bill obligates the Secretary of Defense to develop enforcement protocols and impose penalties for any breaches.
What does this mean for military AI? This legislative action isn’t unprecedented; prior efforts have aimed at establishing regulations governing the Pentagon's use of artificial intelligence. The fiscal year 2026 NDAA contained similar provisions, indicating a continued legislative effort rather than a singular reaction to emerging challenges.
The framework proposed by Senator Gillibrand reflects a strategic approach aimed at safeguarding against catastrophic consequences that could arise from algorithmic errors or unregulated autonomous systems. With nuclear weaponry being one of the highest stakes considerations, its exclusion from AI oversight remains paramount.
The surveillance limitation serves a different purpose. By enshrining this restriction in law, it ensures consistent enforcement regardless of potential shifts in executive branch policies, thereby providing a stronger framework against encroachment on civil liberties.