The World Is Becoming AI-Dependent. Are We Ready?

By Suad Seferi · 14 November 2026 at 12:20

As artificial intelligence becomes embedded in everyday life, recent disruptions expose an uncomfortable question: are the systems supporting this transformation reliable enough? On November 13, the United States temporarily stopped departing flights nationwide after communications failures disrupted air traffic control systems. A day earlier, a major Microsoft outage affected businesses, government institutions and millions of people relying on digital services. Meanwhile, European authorities have been urging citizens to prepare for emergencies by maintaining essential supplies for at least 72 hours. Three developments, each with a different explanation. There is no evidence that they were coordinated or caused by artificial intelligence. Yet I believe they reveal something important about the direction in which our world is moving. We are rapidly transforming society into an environment where almost everything depends on interconnected digital systems. Artificial intelligence is becoming the next layer of that infrastructure, increasingly influencing how businesses operate, how governments make decisions and how people interact with technology. But recent events raise a question that deserves more attention than another AI product launch. Are we building an intelligent future on infrastructure that is not prepared to support it? When technology stops, society stops The disruption in American aviation was extraordinary. The Federal Aviation Administration ordered a nationwide ground stop after telecommunications problems affected radar data and communications with aircraft. The shutdown lasted more than four hours, disrupting thousands of flights. According to the US Department of Transportation, ageing communications equipment was involved. Officials reported that overloaded switches and routers had caused instability, while the investigation continued. The incident was not attributed to AI. In fact, officials specifically excluded the FAA's new AI-assisted forecasting tool from responsibility. But consider what happened. One of the world's most technologically advanced countries was forced to suspend aircraft departures because critical infrastructure could no longer operate with sufficient reliability. The decision to stop flights was a necessary safety precaution. It also demonstrated how dependent a complex physical system has become on its digital foundations. Something similar happened with Microsoft. On November 12, problems affecting Microsoft 365 disrupted Outlook, Teams and other services. The consequences extended beyond Microsoft's ecosystem, affecting integrations used by OpenAI and Anthropic. An AI assistant may remain operational, but if it cannot access the documents, emails or business systems required to complete a task, its usefulness becomes severely limited. That is the problem with technological interdependence. A failure in one component can disrupt services that appear entirely separate to the people using them. And as AI becomes more integrated into these systems, the number of dependencies will increase. Why is Europe preparing for 72 hours without essential services? In March 2025, the European Commission introduced its Preparedness Union Strategy, encouraging citizens to be capable of meeting their essential needs for at least 72 hours during emergencies. On November 3, 2026, the Commission presented additional guidelines covering emergency supplies, preparedness education and crisis communication. The initiative addresses multiple risks, including natural disasters, cyberattacks, geopolitical instability and interruptions to essential services. It is not an announcement of an approaching catastrophe. Nevertheless, I find the underlying message significant. For decades, European societies have been encouraged to digitalise almost every aspect of daily life. Banking, payments, public services, communication and business operations have progressively moved online. Now, European authorities are reminding citizens that they should also be capable of functioning temporarily without some of those systems. In the Netherlands, the central bank and other institutions have recommended keeping emergency cash available in case electronic payments become inaccessible. There is a certain irony here. We are discussing artificial intelligence capable of writing software, analysing complex information and automating business operations, while simultaneously reminding citizens to keep physical money, drinking water and essential supplies at home. This is not evidence that digital transformation has failed. It is an acknowledgement that technological progress does not eliminate vulnerability. AI is becoming infrastructure, not simply another technology The global AI discussion still focuses heavily on models, performance benchmarks, assistants and competition between technology companies. But a more fundamental transformation is taking place. AI is gradually moving from being a tool we choose to use into a capability embedded within systems we depend on. Banks are introducing AI into fraud detection and risk assessment. Healthcare organisations are adopting AI-assisted diagnostics. Companies are automating business processes. Governments are exploring intelligent public services, while defence organisations are expanding AI applications in intelligence, surveillance and operational planning. The International Energy Agency's 2026 analysis projects that global data-centre electricity consumption could rise from approximately 485 terawatt-hours in 2025 to 950 terawatt-hours by 2030. AI is a major driver of that expansion. This is no longer just a software revolution. It requires electricity, telecommunications, data centres, specialised hardware, secure networks and enormous financial investment. We are constructing the physical foundations of an AI-dependent economy. What concerns me is that investment in technological capability can move much faster than investment in the res…