news.today
Hello · Register today | LoginHello · My Account[—] [33]
Tech

US tech outages spark debate over AI vulnerability detection

Summarised from 2 outlets · Updated 19 Nov, 23:38 · Archive

Recent outages affecting airlines, banks and tech platforms across the US have sparked discussion about whether artificial intelligence is finding software vulnerabilities faster than companies can fix them.

Recent outages have disrupted major US systems, including airline booking networks, banking platforms, Apple Pay, NASA's internal communications and Microsoft's Outlook and Teams services. The FAA grounded all flights for several hours, only the third such incident since 9/11.

While authorities have ruled out state-sponsored cyberattacks, cybersecurity experts suggest a different cause. A source told the New York Post that AI systems are finding software vulnerabilities faster than companies can repair them. Modern AI models can identify dozens of flaws in a single review, whereas Microsoft's November security update fixed 1,235 vulnerabilities this year compared to 63 last year.

The challenge is compounded by legacy systems. Airlines, transportation networks and banking platforms often run on outdated software assembled from different systems over decades, making it difficult to process multiple security updates simultaneously. When companies attempt to fix these vulnerabilities, the updates can cause systems to crash. Anthropic and OpenAI warned more than six months ago that their newest models could identify and exploit unknown flaws in major systems without specific guidance to do so.

Companies and US officials have attributed the outages to technical issues rather than deliberate attacks, though the timing of multiple outages within two weeks has drawn comparisons to Y2K-style systemic failures.

US tech outages spark debate over AI vulnerability detection
Image: New York Post

Comments (0)

How it is being reported

In this story: Apple Pay · NASA · FAA · Microsoft · OpenAI · Anthropic · United States

This summary was written by AI from the headlines and standfirsts above, and states as fact only what at least two outlets report. How we use AI · Report a problem · Corrections log