A few hours ago, I attempted to rent a Lime bike to reach my therapy session. The app froze and closed after I authorised payment via Apple Pay. Several subsequent attempts failed in the same way. I tried Uber, which also refused payment. A check of DownDownDetector confirmed a widespread failure affecting Apple Pay, the App Store, and other services. I left my phone in my pocket and started jogging.
A system in decline
Being unable to rent a bike for a few minutes is a minor inconvenience. Yet the mental effort required to figure out what was broken has become a weekly routine for many people. Over the weekend, contacts outside my industry asked about the grounding of flights, bank outages, and Microsoft Teams failures.
The default assumption is that these events are caused by cyberattacks or rogue AI agents. This fear often overshadows the reality: the systems are failing because they are poorly maintained. It is not a single hack, but a dependency chain collapsing under its own weight.
While AI and cyberattacks are factors, they are symptoms of broader systemic issues. We are seeing infrastructure failures that reflect trends across society, government, and the tech sector over the last few years.
Why things are breaking
When outages occur, there is a tendency to link them into a single narrative. Incidents that previously went unnoticed suddenly become part of the story. For example, a cyberattack on the International Ice Hockey Federation has received significant attention despite being an isolated event.
The root causes for this trend include:
- Mass layoffs of tech workers and software developers
- Widespread deployment of AI-assisted products and software updates
- Reductions in government expertise and oversight
- Loss of institutional knowledge due to hiring cuts
- A lack of investment in upgrading infrastructure and redundancy
- Government contracts awarded to political allies
- Use of AI and LLMs to find exploits and write code
- Rushed patches generated by AI in response to security threats
When Elon Musk renamed Twitter to X, many predicted the site would fail due to a lack of expertise. While X has not collapsed yet, this philosophy of firing experts and relying on AI is now the standard strategy for the economy and government.
We are seeing a mix of clustered attacks, outages, and errors that occur when companies move too fast or lack the knowledge to manage complexity. This approach treats critical systems as if they can be duct-taped together without long-term consequences.
“We are Ship of Theseus-ing our entire economy and infrastructure”
We know this is happening. Following the PHP2Shell exploit, which was discovered with the help of large language models, numerous websites went down as they rushed to push updates. The time between a vulnerability being found and exploited is shorter than ever.
As I wrote this, a California company announced it had used LLMs to find serious vulnerabilities in Java. Companies and agencies are pushing updates on live systems, including air traffic control, partly because of the perceived threat of AI. The National Security Agency reported earlier Tuesday that internal systems went down due to temporary degradation during scheduled maintenance.
Nick Andersen, acting director of the Cybersecurity and Infrastructure Security Agency, said resilience must be built into our infrastructure. He noted that building redundancy requires people who understand how systems fail and how to plan for it.
“Ultimately being able to prevail by being able to demonstrate resilience against disruptions, whether they occur because of things like an IT outage or a bad patch, or a modernization update gone awry, or because of a malicious cyber incident — we want to make sure we’re engineering in the resilience that is going to be able to anticipate for and ultimately overcome those outages and those impacts,” Andersen said at an event called FedTalks.
Eventually, expertise in complex systems may become a priority again. However, the last few years have seen companies and governments move quickly to reduce reliance on humans and offload work to AI. This shift aims to save on labour costs or advance an anti-humanist political project.
A software developer named Jimmy Breck-McKye wrote a piece titled “Software is breaking down at a systemic level.” He references Kristen Maevyn’s “The Decoherence Trap,” which argues that AI-written code is creating a situation where people no longer understand their own products.
“Most software engineers now are unable to tackle engineering problems without some form of AI assistance. That’s not because they’re incompetent — they just can’t keep on top of the complexity,” Breck-McKye writes. “Even if they could once, they now face a tsunami of AI generated code. This is affecting everyone, at all levels — we are all fighting the tide of small, incoherent changes that look rational in isolation but slowly lose sense as a whole.”
People increasingly do not understand how their code works. Speed is prioritising over stability, and bad things are happening. Individual outages may be coincidental, but many are linked to a broader lack of investment in infrastructure and expertise. We are replacing our entire economy and infrastructure piece by piece, so we must expect failures to continue.




