Cybersecurity in the Skies: AI in Commercial Aviation and the Imperative of Human Oversight
Navigating the integration of new technologies, such as AI in aviation, and recent events affecting the airline transport industry suggest how crucial it is to find a balance between technological innovation and human expertise.
While AI will inevitably contribute to efficiency and automation, the human element remains indispensable for assessing complex situations, responding to unforeseen challenges, and ensuring the safety and security of air travel.
Collaboration between emerging technologies and human pilots appears as a strategic imperative, fortifying aviation safety against new hazards, such as cyber threats. This balance will become the cornerstone of resilient and secure aviation as we look to the future skies.
These days, in many areas of the world, the aviation system is experiencing an important system deficiency: if the GPS signals are jammed or spoofed, aircraft have issues detecting their position and, at the same time, trigger uncontrolled warnings.
GNSS jamming and spoofing occur when aircraft GNSS signals experience interference from stronger radio signals.
Jamming is intentional Radio Frequency Interference (RFI) with GNSS signals (EASA, 2022) that causes a loss of position precision. Spoofing, on the other hand, is an intelligent form of interference that tricks aircraft systems into believing to be in one position rather than another (Lopez & Simsky, 2021). During a spoofing attack (EASA, 2024), a nearby radio transmitter sends fake signals to the target receiver.
EASA Acting Executive Director Tytgat (2024) says: "We have seen a sharp rise in attacks on these systems, which poses a safety risk. EASA is tackling the risk specific to these new technologies".
Zee (2023) argues, "The industry has been slow to come to terms with the issue, leaving flight crews alone to find ways of detecting and mitigating GPS spoofing".
EASA (2024) recognises that such occurrences belong to the domain of cybersecurity and identifies the crucial role of the flight crew in mitigating their potential effects by identifying the immediate "need to ensure that pilots and crews can identify the risks and know how to react and land safely".
These recent events make us reflect on the future of commercial aviation, especially with the introduction of new cutting-edge technologies like artificial intelligence. AI has the potential to revolutionise aviation and create advanced automation. However, all this raises important questions and challenges that need to be addressed.
In these evolving scenarios, the recent incidents of GPS spoofing and interference serve as a reminder of the need for a comprehensive approach to aviation safety that integrates both new technologies and human oversight.
A human-centred approach to innovation
Integrating artificial intelligence in commercial aircraft holds great promise for efficiency. However, the increasing reliance on complex software systems raises concerns about emerging threats, such as cybersecurity threats.
As evidence of the novelty of this phenomenon, we can consider that as of today, the consolidated versions of many airlines' operations manuals provide exhaustive and detailed instructions regarding physical security threats such as bomb threats, unruly passengers, or hijacking, but still little about this new type of cyber hazard, which is generally addressed through bulletins or in the FCOMs.
At the same time, the possibility of seeing AI-piloted aircraft is becoming more than a hypothesis. So, reflecting on recent events, it is becoming increasingly important to question the vulnerability of aircraft piloted solely by AI to hacking attacks and security breaches to define viable strategies and defences. In addressing the recent GPS events, the human element emerges as a crucial and essential countermeasure against unforeseen threats, such as cybersecurity events.
Regulatory bodies and entities, such as EASA, view pilots as a necessary line of defence against threats like GPS jamming and spoofing.
EASA SIB No.: 2022-02R2 outlines steps airlines should take, including:
• Training pilots to identify and report GNSS issues.
• Educating pilots on jamming and spoofing scenarios during recurrent training.
• Equipping pilots with skills to react to these threats, e.g. by using alternative navigation methods.
By issuing this document, the regulator assigns pilots a crucial role in safeguarding safety and recognises their critical role in facing evolving threats.
The surge in GPS-related disruptions emphasises some limitations of technology in addressing unforeseen, subtle challenges. Drawing on their experience and ability to evaluate contextual information, human pilots play a critical role in identifying anomalies, cross-referencing data, and making informed decisions to ensure flight safety.
Human pilots bring unique skills and competencies into the cockpit, including the ability to respond to unforeseen circumstances, adapt to dynamic situations, and make critical decisions in the interest of passenger safety.
In the face of security events, a human pilot's immediate intervention and decision-making capabilities become invaluable. While AI continues to advance, the complex nature of human cognition ensures a resilient defence mechanism against potential threats.
As technology evolves, a human-centric approach to AI in aviation (EASA, 2023) and the collaboration between AI and human expertise are imperative to fortify aviation safety against emerging threats.
References
EASA, (2023). SIB No.: 2022-02R2. European Aviation Safety Agency. Cologne, Germany.
EASA (2024). EASA partners with IATA to counter aviation safety threats from GNSS spoofing and jamming. EASA Press releases. European Aviation Safety Agency. Cologne, Germany.
EASA (2023). Artificial Intelligence Roadmap 2.0. European Aviation Safety Agency. Cologne, Germany.
ICAO (2022). Aviation Security Manual, DOC. 8973. International Civil Aviation Organization, Montreal, Quebec, Canada.
Lopez, G., Simsky, M. (2021). What is GNSS Spoofing? GIM International.
Zee, M. (2023). GPS Spoofing Update: Map, Scenarios and Guidance. OPS Group.
ENISA (2021). Cybersecurity Challenges in the Uptake of Artificial Intelligence in Autonomous Driving. ENISA Press releases.