AI Drives New Opportunities and Risks in Space
AI has emerged as a transformative force in the space sector, reshaping the design, operation, and governance of activities in orbit. While this technology opens new possibilities for innovation, it also exacerbates existing challenges such as market concentration and cybersecurity.
The Accelerating Trends Shaping AI in Space
The number of satellites currently in orbit is approximately 15,000, with projections estimating this could rise to 100,000 by 2030. If all current filings for planned low Earth orbit (LEO) satellites at the Federal Communications Commission (FCC) are launched, this number could balloon to half a million by the end of the 2030s. This exponential growth is primarily driven by commercial expansion in mega-constellations, notably Starlink, which currently operates about two-thirds of all active satellites, surpassing the total of all countries combined.
The surge in satellites is accompanied by an explosion of data. For instance, NASA’s Earth Observation System Data and Information System archive has already reached 100 petabytes (PB) of data, expected to grow to 320 PB by 2030. Concurrently, stakeholders access 5 PB of data monthly from programs like the National Oceanic and Atmospheric Administration (NOAA)’s Open Data Dissemination (NODD).
In this landscape, the true value for future companies, governments, scientists, and innovators lies in the “time-to-insight” derived from this data. Edge AI, implemented directly on satellites along with cloud-to-edge AI on the ground, will be essential for managing escalating data volumes. Edge computing allows for real-time processing aboard spacecraft, crucial given the communication delays in deep space. This capability enables satellites to filter and prioritize data before transmission, reducing bandwidth requirements and facilitating autonomous decision-making.
Opportunities
Value-Add in the Space Industry
Experts like Raycho Raychev, founder of EnduroSat, argue that as AI models become more commoditized, unique data and service provision will emerge as the most valuable resources. Space companies, with their vast data access and capabilities in big data analytics, are ideally positioned to leverage this shift.
Integrating AI to Unlock New Use Cases
Recent advancements, such as those from Stanford researchers, demonstrate successful AI integration into existing technologies, enhancing performance and creating new applications. For example, AI-enabled computer vision is crucial for missions like NASA’s 2020 Mars mission, which successfully landed on previously hazardous terrains. The Ingenuity Mars Helicopter has autonomously performed 72 flights over Martian landscapes, with aspirations to land in even more challenging conditions.
Challenges
Despite its potential, the integration of AI in the space sector introduces complex challenges related to both business and governance.
Scaling and Vertical Integration
Raychev and Mason highlight that vertical integration can hinder scaling for space companies. Currently, these companies manage multiple functions—from infrastructure development to data analytics—stretching their resources thin and complicating growth. A more collaborative approach, where some players focus on infrastructure while others specialize in intelligence applications, could alleviate this issue.
Market Concentration
The integration of AI also poses significant risks for governments, as it may lead to further market concentration among a few capital-intensive firms. This concentration can marginalize countries lacking access to critical Earth observation data and AI algorithms, potentially compromising their ability to make informed decisions in areas like agriculture and national security.
Governance and Cybersecurity
Both the AI and space industries face intricate cybersecurity and governance challenges that are only magnified by their convergence. With commercial satellites increasingly supporting military intelligence, the risk of cyberattacks has escalated. Incidents include GPS jamming in Europe and ransomware attacks affecting multiple organizations in 2024.
The dual-use nature of both AI and space technologies compounds these risks, creating new categories of threats that existing governance frameworks are ill-equipped to manage. AI-driven space-based decisions can occur in microseconds, challenging traditional governance models that rely on human oversight.
Strategies for Governance
As satellite launches are expected to accelerate, prioritizing safety in both public and private sectors is essential. Recommendations from the United Nations Office for Outer Space Affairs (UNOOSA) advocate for human-in-the-loop systems for low-latency operations and human-on-the-loop frameworks with safeguards for deep-space missions.
UNOOSA encourages the establishment of new technical standards, such as explainable AI for space-grade hardware, and the incorporation of decision logs in the private sector. Governments are urged to develop an international code of practice for AI in space.
Moreover, the public sector must adapt to the evolving landscape where commercial satellites play a crucial role in military strategies, exemplified by the U.S. Department of Defense’s Commercial Space Integration Strategy.
To address market concentration, UNOOSA recommends fostering international collaborations in AI-for-Earth observation projects and ensuring public data from funded missions is accessible.
Finally, developing a global cybersecurity protocol for space will be critical. Experts advocate for real-time information sharing and coordinated responses to threats. Integrating AI innovations can help detect potential cybersecurity risks and mitigate physical threats despite communication delays.
Conclusion
The challenges and opportunities presented by AI are set to reshape the future of the space industry. Coordinated yet agile governance will be vital for the success of commercial ventures and governmental initiatives, enabling advancements that improve quality of life on Earth and beyond.
