For decades, aviation safety risk management has relied on building layers of defense against human error, operational hazards, or mechanical failures. However, today we face threats that go far beyond these well-known areas of risk. These new emerging risks—although directly affecting safety—are still managed within their own siloed risk management systems.
We are talking about emerging risks such as conflict zones impacting routing decisions, health crises like COVID-19 disrupting the global network, drone interference in controlled airspace, socio-economic factors that could weaken safety oversight, and environmental risks such as shifting wind patterns with direct impact on flight safety. The aviation sector is now confronted with a much broader safety landscape. These risks do not occur in isolation. A conflict zone closure can trigger rerouting, exposing new cyber vulnerabilities in communication networks, or altering flight patterns with environmental implications.
Many of these threats are captured in the agendas of national and international authorities and are closely monitored by operational stakeholders. Yet, as of today, the system lacks the capacity to build a holistic or integrated picture of risk across safety.
Among all these risks, cybersecurity stands out as one of the most pressing. In recent years, we constantly hear about it: GNSS jamming across Europe forcing diversions, ransomware and DDoS attacks on navigation and/or airport systems. From a deliberately falsified maintenance record to a compromised ATC subsystem, or—as in a current example—a poisoned AI training dataset, these are events that go beyond traditional hazards but can escalate into operational disruption and, in the worst case, accidents.
In the case of cybersecurity attacks, the challenge is amplified by the very nature of the sector, which is a true system of systems. Aviation relies on a network of interconnected subsystems (aircraft, ground infrastructure, supply chains, providers), with multiple dependencies—each node becoming a potential vulnerability: the supplier’s software, the airport’s network, the data link itself. Jamming or spoofing events, for example, directly affect GPS signals, which in turn can cause conflicts in minimum aircraft separation and thus act as de facto precursors of a mid-air collision (see bowtie diagram below).
From all this, one common lesson emerges: aviation can no longer manage risk in silos. Emerging threats—whether cyber, environmental, or geopolitical—interact with each other, generating risk scenarios that current systems are not capable of capturing.
Where the sector stands today: fragmented frameworks and reactive posture
Today, the need for an integrated approach to risk management is globally acknowledged. IATA’s Integrated Risk and Resilience Management (IRRM) manual explicitly calls for the integration of Safety (SMS), Quality (QMS), Security (SeMS), and Emergency Response (ERP) under one unified framework. ICAO’s Safety Management Manual (Doc 9859) also stresses the importance of harmonising safety and security risk management, while EASA regulations allow for a single management system that can include not only safety and security, but also environment and occupational safety.
With respect to cyber risks, Europe has also introduced mechanisms for their integrated management. EASA’s Part-IS regulation has created binding obligations for managing information security risks with a safety impact.
Despite the existence of these frameworks, achieving effective integration in practice remains complex. Safety events and cyber incidents, for instance, are often recorded in separate taxonomies, with risk registers stored in different systems, and therefore difficult to correlate. For this reason, even though we know that jamming and spoofing events impacting GPS systems can escalate into separation minima infringements, it is not straightforward to evaluate at scale the level of exposure to these events, or how often and how severely a cyber precursor impacts flight safety.
The current challenge is not the framework itself, but the need to establish intelligence capabilities that enable truly integrated risk management. Many organisations may be compliant on paper, with systems monitoring risks within their domains, but without real integration of risk management in practice. Without integration, there is no resilience—leaving the sector dependent on reactive management of these risks.
How Data & AI can support integrated risk management in aviation
Strengthening integrated risk management in aviation requires more than frameworks and compliance—it demands building ‘safety intelligence’ capabilities. Data and AI play a central role in this transformation for three key reasons:
- To build a global picture of the ‘visible’ risk: The first step is to understand the tip of the iceberg. GNSS spoofing may appear in cybersecurity logs and in safety incident reports, but linking them to generate a unified picture of risk is challenging. Reports may not be directly comparable due to different taxonomies, or they may span time periods that do not align. Advanced analytics and AI tools can help overcome these gaps, creating a unified view across logs and datasets.
- To discover hidden vulnerabilities before they escalate: The most dangerous risks are often those not yet modelled. Frontline operators may not realise that the safety event they are recording could have been induced by a cyber incident. In some cases, cyber-attacks may create operational disruptions or degrade safety margins without being reported, but still capable of escalation. AI can help uncover these blind spots by detecting patterns of co-occurrence across domains and quantifying vulnerabilities that conventional taxonomies cannot.
- To quantify operational and financial disruption: Beyond identifying risks, it is equally important to assess their impact. Data & AI can connect anomalies to measurable outcomes—such as flight delays, diversions, supply chain disruptions, or cascading passenger impacts—translating them into operational and financial terms that decision-makers can act upon. This not only strengthens resilience but also informs better resource allocation and investment in mitigation strategies.
ALG has already explored this approach in practice. By combining cyber event databases with safety occurrence reports, we tested how emerging risks could be correlated across domains. At first glance, correlations were limited, GNSS jamming and spoofing events being the most obvious. But when analysed through graph-theory methods and machine-learning models, clusters emerged that pointed to “induced events” otherwise invisible to standard reporting.
Furthermore, integrating LLMs into the safety intelligence process enabled contextual detection of cyber-related risks within safety reports at scale, without proper actual identification by the front-line reporter. This work provided us with a new lens, a way to surface unknown safety dependencies and quantify vulnerabilities.
The call to action: building resilience through data-driven integrated risk management & collaboration
The aviation sector is facing a new safety landscape with multiple emerging risks that require building a broader and more integrated vision. Among these, cyberattacks stand out as a fast-growing and systemic threat.
However, this situation is not entirely new for aviation. When accident rates peaked in the 1970s, the response was not to add more checklists but to embed a new culture of safety management. The challenge today is similar, requiring another shift in the way the industry manages risk and the tools used to that end.
Conceptual frameworks for integrated risk management already exist, but as always, the challenge lies in implementation. This is where Data and AI tools can make a difference. Integrated data collection and analysis platforms can merge safety, cyber, and environmental data into a single view. Through advanced analytics and AI models, cross-correlations across domains can be uncovered, as well as hidden vulnerabilities and previously unknown risks. Additionally, Agentic AI can support interaction and decision-making in the new integrated governance to be defined.
The recommendation for industry leaders is straightforward: soon, compliance alone will no longer be enough. The sector must start developing intelligence capabilities that make it more resilient to emerging risks, maintain safety levels, safeguard operational excellence, and protect organisational assets.
Finally, collaboration will be essential. Regional and international collaborative programs—such as EASA’s Data4Safety, where ALG participates as Data Platform and Analytics provider—will accelerate this journey. They will enable data collection beyond organisational and domain silos, and foster the development and democratization of safety intelligence across the sector, ultimately helping to build a global, integrated picture of risk.
About the authors
Antonio Cabeza
Aeronautical engineer with a MSc in Big Data and Advanced analytics. Specialises in strategical projects in airport and air traffic operations with wide expertise in digitalisation and regulatory related projects. Senior Engagement Manager at ALG.
Núria Alsina Pujol
Aeronautical engineer with specialisation in air navigation and systems, certified in project management and scrum methodology. Principal at ALG and head of digitalisation and advanced analytics in the Transportation practice.
Mar Villalonga Torres
Aeronautical Engineer with an MSc in Autonomous Vehicles Dynamics and Control. Experienced in developing and applying advanced data analysis and modelling techniques in the field of aviation safety and operations. Consultant at the AI and Digital Transformation at ALG.
