Rapid advances in artificial intelligence (AI) now pose an imminent threat to satellite systems, potentially triggering a catastrophic cascade of collisions in orbit within the next two years. Cybersecurity experts warn that AI-driven attacks could hijack spacecraft, turning them into uncontrollable weapons and jeopardizing the safety of the entire space environment.
The Evolving Threat Landscape
For decades, the space sector has largely ignored cybersecurity risks, leaving thousands of older, vulnerable satellites exposed. Now, the emergence of “agentic AI” – autonomous systems powered by Large Language Models (LLMs) like ChatGPT and Gemini – is drastically lowering the barrier for attackers. LLMs allow even those without specialized knowledge to quickly understand spacecraft operations and identify vulnerabilities.
AI’s acceleration is critical. State-sponsored hackers, such as Russia’s Fancy Bear, have already begun using LLMs to gather intelligence on space systems. This drastically reduces the time needed to exploit existing weaknesses. As Andrzej Olchawa, a space cybersecurity engineer, points out, “LLMs have drastically lowered the barrier to understanding spacecraft operations… adversaries with no prior knowledge can now cause real harm.”
How an Attack Could Unfold
The most immediate danger is the potential for AI-driven attacks to hijack satellites and force collisions. Kristjan Keskküla, Head of Space Cyber Range at the CR14 cybersecurity center in Estonia, explains that affecting just one satellite can create chaos, especially given the current density of over 8,000 satellites in low Earth orbit.
A deliberate crash could generate thousands of debris fragments, rendering orbital space unsafe for years. Attack methods include jamming, spoofing communication links, or taking complete control of a spacecraft.
The Urgency of Preparation
The space industry is belatedly waking up to these threats, but many systems remain unprotected. Cybersecurity researchers are racing to identify vulnerabilities before attackers exploit them, but AI tools also empower adversaries to discover those holes faster.
CR14, a leading cybersecurity research center, conducts exercises simulating these attacks in virtual environments. Their findings demonstrate that the sophistication and volume of cyberattacks are escalating rapidly. Ukraine, for example, faces “thousands of cyberattacks” daily on critical infrastructure, including satellite communications.
The problem is not just the existence of threats, but their increasing intelligence. Phishing attacks in Europe have risen by 500% in recent years, with AI making them more targeted and convincing. This means attackers can gather more personal information and impersonate trusted contacts to steal sensitive data.
The rapid development of AI has created a dangerous window of vulnerability. The space sector must adapt quickly to defend against this evolving threat, or risk a catastrophic orbital apocalypse.
