ON Thursday (early Friday morning in Manila), hundreds of thousands of users of Microsoft applications, among them Outlook, Teams, SharePoint and Copilot, reported them being offline. Initially, the United States seemed to be the most affected by the sudden outage, but it quickly became obvious that it was a global problem. As of the start of the workday on Friday, one very large Manila-based institution we contacted — which has several thousand employees here, and thousands more working remotely or in offices in several dozen countries — found itself essentially dead in the water, as it relies heavily on Microsoft enterprise systems.
Microsoft announced that it was “aggressively” acting to determine the problem and to take whatever steps were necessary to restore services, which is understandable, as the company was likely hemorrhaging millions of dollars a minute while its systems were offline. The company attributed the breakdown to an overload of traffic, and said it was working to reroute some of the load and bring additional server resources online to restore normal operations. Several wags on social media commented: “Microsoft is trying the ‘turn it off and then turn it back on again’ fix.”
All fun aside, the concerning issue in all of this is not really about Microsoft’s systems going down, disruptive and costly though that may have been for some. Rather, the incident is a reminder of how vulnerable our modern world is to being brought to a standstill by relatively minor technical failures or in some potential cases, hostile action against a small part of the global infrastructure. The systems we rely on to conduct our business and communicate are only increasing in fragility as the shift toward greater use of “cloud” platforms and growing integration of artificial intelligence tools accelerates.
The Microsoft outage should be a stark reminder to everyone that reliance on a single platform is a serious risk. The enterprise packages from companies such as Microsoft or Google are ordinarily quite useful and reliable, but they are not infallible. Breakdowns may be rare, but can be disastrous for your business or institution when they occur. The large organization we mentioned above is a good example; as long as the outage persists, most of its personnel are unable to communicate with each other via email, or the Teams online meeting program, and critical, time-dependent documents that must be accessed by multiple departments are out of reach.
The vulnerability goes beyond software systems as well, particularly for an archipelago like the Philippines. We are connected to the rest of the world, and our various islands are connected to each other, through a number of undersea cables; there are at present 18 international connections and a number of domestic cables. These are vulnerable to physical interruption — earthquakes, accidents such as being cut or snagged by ship anchors, or deliberate acts of sabotage by hostile actors.
There have been examples of the first two in recent years affecting the Philippines, but fortunately not the latter. But incidents of that kind in northern Europe early in the Russia-Ukraine war demonstrated how easily that kind of attack could be carried out. These cables are becoming increasingly congested as the world races toward a state of “everything online all the time,” such that the temporary loss of even one or two of them can create bottlenecks that can interrupt business activities and communications. And despite their marketing, satellite internet networks such as Starlink are not a consequential alternative. Even if the Starlink orbital network, for example, was several orders of magnitude larger than it is, the amount of internet traffic it could manage would still be a fraction of what the cable network can carry.
No business or institution should be forced to stop working completely because one’s business applications provider suffered a server breakdown, or an errant captain snapped an undersea internet cable with his ship’s anchor. Every organization should review its internet-based activities and assess their vulnerability, particularly for “mission-critical” tasks or processes. Internet-based enterprise platforms are wonderfully versatile and efficient, until they are not, as the millions of clients of Microsoft’s suite of online tools discovered at the end of this week. Having a backup plan and alternative channels through which your most important work can be done is a must.
