The Squid’s InkSquidGrid’s weekly read of the internet — fresh ink every Monday

Week 40, 2026 · Sep 21 – Sep 27

Covers Monday Sep 21 00:00 to Monday Sep 28 00:00 UTC.

How to read this: we measure millions of internet routes around the world. Every speed number below is how many milliseconds (thousandths of a second) slower a route ran than its own normal speed, so higher is worse. Each section says whether things got better or worse than last week.

2,224,322internet routes we measure
687,268,576measurements taken this week down 5% from last week
1,539outages and problems companies reported down 9% from last week
512of them rated major or critical

Bharti Airtel Limited’s rough weekthe week’s story

Better than last week. Bharti Airtel Limited is one of the large networks that carry internet traffic between countries. Routes through it ran 31.5 milliseconds slower than normal on average this week (last week: 32.6). 1,734 routes through it were badly slowed (more than 25 milliseconds over normal) at some point. The chart shows how many, hour by hour:

Signed-in customers see the affected routes, the carrier on each side of every changed handoff, and the regional split on the Global fleet view.

The networks that slowed down the most

Large carrier networks, ranked by how much slower than normal their routes ran this week. Big networks get no allowance for their size.

Networkthis week, hour by hourslower than normalroutes badly slowedlast weekcompared with last week
SG.GS (Singapore network operator)46.8 ms82946.4 msAbout the same as last week
FLAG Telecom36.4 ms27331.0 msWorse than last week
Bharti Airtel Limited31.5 ms1,73432.6 msBetter than last week
SEACOM26.5 ms42226.3 msAbout the same as last week
TransTelecom26.4 ms53226.7 msAbout the same as last week
Gateway Communications (Belgium)24.9 ms2,39226.1 msBetter than last week

“Routes badly slowed” counts routes that ran more than 25 milliseconds over normal at some point this week.

1,539 outages and problems reported by companies

Fewer than last week (down 9%). These are problems online services posted on their own status pages. Last week: 1,697.

SaaS948
Payments & fintech225
Crypto55
AI53
Gaming46
Education & research42
Cloud & hosting42
CDN & edge37

Online business software (SaaS) had the most problems again, as it usually does.

What tended to break, in providers’ own words: connectivity ×140 · slowness ×93 · auth & login ×78 · elevated errors ×66

Longest problems to fix: University of Nebraska–Lincoln — GlobalProtect VPN under investigation (486 hours, about 20 days); Bit2Me — Aviso de deslistado de MemeCore (M) / Delisting Notice — MemeCore (M) (479 hours, about 20 days); Bit2Me — Aviso de deslistado de Covalent X Token (CXT) / Delisting Notice — Cov (479 hours, about 20 days).

Ground eventsthe physical world, and whether the internet flinched

2026-09-25 AT&T core network stumble felt in the US, UK and Singapore
Early on Thursday, Sept 25, AT&T had a short outage on its core network. Monitoring firm ThousandEyes reported it, and Network World published the report. AT&T is one of the big US carriers that other networks rely on to carry their traffic. The trouble started around 05:50 UTC at AT&T sites in Chicago. It seemed to clear, then came back minutes later in Chicago, Ashburn (Virginia) and Dallas. Other providers and their customers that depend on AT&T were hit in the US, the UK and Singapore. The outage came and went over a short period. No cause was given in the report.
Network World outage report
2026-09-22 Cloudflare outage hits users in the US and the Philippines
On Monday, Sept 22, Cloudflare had an outage, ThousandEyes reported via Network World. Many websites use Cloudflare to deliver their pages quickly and to protect them from attacks. The problem began around 13:38 UTC and was centred on Cloudflare's site in Los Angeles. Providers and customers in the US and the Philippines were affected. The outage was short. No cause was given in the report.
Network World outage report

Ongoing, week 15 of tracking: Iran, the same as last week. People are back online after the months-long shutdown that ended in May, but access is still heavily restricted. A new bill to tighten control of the internet was put to lawmakers on Aug 26. Radio Farda (RFE/RL) reported it was unclear when it would be voted on. Routes into Iran (60 we measure) ran 0.7 milliseconds slower than normal, the same as last week.

We cover earthquakes of magnitude 7.0 or more and storms at the top two disaster alert levels. “Our fleet’s view” is what our own measurements saw, not a damage report.

Subsea watchwhere the internet is wet

For the feedpaste-ready

The Squid's Ink -- Week 40 One stumble, three countries. Early on Thursday, Sept 25, AT&T had a short outage on its core network, according to monitoring firm ThousandEyes, reported by Network World. The trouble started at AT&T sites in Chicago. It seemed to clear, then came back minutes later in Chicago, Virginia and Texas. Why did people far from the US notice? AT&T is one of the big carriers that other networks rely on to move their traffic. When it stumbled, providers and customers in the UK and Singapore felt it as well as those in the US. The lesson: the internet has no borders inside it. A short problem at one large carrier can show up an ocean away. Elsewhere: Iran was the same as last week on our measurements. Access there is still heavily restricted. #InternetHealth #TheSquidsInk #SquidGrid

Every figure derives from SquidGrid’s own measurement fleet and its archive of provider-published incidents; provider self-reports are labeled as such. Provider Assessment Reports carry the per-provider detail. The internet was mostly fine this week. We measured anyway.