AI Drives New Opportunities and Risks in Space

Researchers at Stanford brought machine learning to robots aboard the International Space Station in 2025, enabling them to plan movements 50 percent to 60 percent faster. This example shows artificial intelligence moving to the center of space sector activities, including how spacecraft are designed, operated and governed. The space economy hit a record 613 billion dollars in value in 2024, with 78 percent in the commercial sector, according to McKinsey. The firm estimates the sector could reach 1.8 trillion dollars by 2035, driven by petabyte-scale data, mega-constellations and operations that exceed human decision capacity. Approximately 15,000 satellites are now in orbit, with projections reaching 100,000 by 2030 and potentially half a million by the end of the 2030s if all planned low Earth orbit filings launch. Starlink accounts for about two-thirds of active satellites. NASA's Earth Observation System Data and Information System archive holds 100 petabytes of data and is expected to reach 320 petabytes by 2030, while the National Oceanic and Atmospheric Administration's Open Data Dissemination program handles 5 petabytes per month and Europe's Copernicus program exceeds 78 petabytes. Edge AI on satellites and cloud-to-edge systems on the ground support real-time processing and data prioritization, which reduces bandwidth needs and enables autonomous decisions, especially where communication delays range from minutes to hours. Raychev, founder of EnduroSat, and Mason, Chief Space Officer at Planet, said that as AI models become commoditized, unique data and service provision will likely become the scarcest resources. AI-enabled computer vision and terrain-relative navigation have supported missions such as NASA's 2020 Mars landing and the Ingenuity helicopter's 72 flights. Vertical integration remains prohibitive for scaling, as companies handle infrastructure, data, analytics and AI capabilities themselves. Market concentration in a few capital-intensive firms raises concerns about exclusion and dependency for countries without access to the same data or platforms. Cybersecurity incidents include GPS jamming in Europe, attacks on space agencies in Japan and Poland, and ransomware against 25 space-sector organizations in 2024. The convergence of AI and space creates new dual-use risks that outpace traditional governance, as AI-driven decisions occur in microseconds. The United Nations Office for Outer Space Affairs recommends human-in-the-loop approaches for low-latency operations and human-on-the-loop safeguards for deep-space missions, along with explainable AI standards, decision logs and an international code of practice. The U.S. Department of Defense issued a Commercial Space Integration Strategy to address commercial systems in military contexts.