tech
Tech Outages Are Piling Up, AP Finds, as AI Raises the Stakes

A new AP News report is asking a straightforward question: as outages at major tech providers keep happening, is the disruption getting worse because more of the technology people rely on now runs through artificial intelligence systems? The AP report frames the pattern as a question for both companies and regular users: is this the new normal, or a phase that better infrastructure can fix?
What Is AP Reporting About the Pace of Outages?
AP News describes tech outages as "piling up," a phrase that signals frequency rather than a single incident. The report does not point to one company or one date — it treats the trend as cumulative, building the case over a string of disruptions rather than a single event. That framing matters for readers trying to judge whether this is a blip or a pattern worth planning around.
Why Does AI Make an Outage Hurt More?
The AP report ties rising stakes directly to AI's growing role in everyday software. When a product adds an AI feature — a chatbot, an image generator, a recommendation engine — it usually means plugging into a cloud-based AI system running somewhere else. If that system goes down, the AI feature goes down with it, and so does whatever else in the product depended on it. The more AI gets built into ordinary apps, the more a single back-end failure can ripple into unrelated services.
Is That AI Footprint Actually Growing?
Separate coverage from HTT News shows the expansion AP is describing. Meta has been rolling out new AI-focused subscription plans, pushing more paying users onto AI-powered tiers of its products. In San Francisco, a creative project described by HTT News runs on an AI system modeled on a fly's brain, another example of AI moving from research labs into consumer-facing tools. Each addition is another dependency that has to stay online for the product to work.
What Changes for Businesses and Users Starting Monday?
AP's framing leaves readers with a practical takeaway rather than a technical one: anyone running a business on AI-dependent software should know what happens if that AI layer disappears for an hour. That means knowing whether core functions — billing, scheduling, customer support — still work without the AI feature turned on, and whether a provider's status page gets checked before a help desk ticket gets filed. For individual users, it means treating AI features as an add-on rather than the foundation, since a provider's AI layer can fail even when the rest of the service holds up.
AP's underlying question — whether frequent outages are becoming the new normal — does not have a settled answer in the reporting. What is clear is that each new AI feature added to a widely used product adds one more point of failure, and one more reason an outage that used to be a minor annoyance can now take more of a product down with it.
For a Bay Area-baked gift, Stirred, Not Shaken ships in the U.S.