Digital dependence can disrupt society without AI or hackers
Monday 16th November 2026 on 09:30 in
Estonia
Recent disruptions to air travel and hospital networks show that digital crises can shake everyday life without artificial intelligence or hackers, R2 commentator Kristjan Port writes in an article published by ERR.
Paul Christiano, a board member of the OpenAI fund, warned during a lecture series honouring Nobel laureate Geoffrey Hinton that, in his view, artificial intelligence could gain control over society within three years. He said significant risks could emerge in the next six to 12 months. His remarks might leave listeners thinking that cancer or a heart attack was unlikely to kill them, while an AI-related failure could do so first.
Christiano said stopping developments already under way would be as futile as a crowd taking on the US military. Aggressive resistance, he argued, would only end in deaths and environmental destruction. The more hopeful message, Port writes, was that people needed to act more wisely. Christiano stressed that his estimate was not a precise calculation but a coin toss.
Even if the coin lands on the more favourable side, society faces challenges unrelated to AI. At the end of last week, air traffic across the United States was halted for a couple of days. The US transport authority said nationwide air traffic stopped on November 13 because of telecommunications disruptions. The cause was not a terrorist attack like the September 11 attacks, but a more ordinary technical incident.
A router installed during a switch from copper cables to a fibre network tried to connect too frequently, disrupting other parts of the network. But the router itself did not stop air traffic. People did, after they were left uncertain whether to continue operating with unreliable communications and radar information or ground thousands of aircraft. They chose to halt flights.
A disruption to Microsoft services affecting more than 400 million users worldwide occurred around the same time. The CrowdStrike incident in July 2024 had already provided a warning: a faulty software update disrupted airlines, hospitals, banks and many other organisations, even though they had done nothing wrong.
Whether caused by hackers or technological complexity, such incidents expose a stark reality, Port writes. Technological progress tends to increase complexity and make each new problem more difficult. In the past, a malicious hack could be blocked or isolated while the system continued to function. More recently, the response has increasingly been to shut down the entire system. The problem is not so much the technical failure itself as the declining ability of system owners to keep operating when something goes wrong.