The disruption is widespread and has come at a busy time for developers. GitHub has conceded that performance is not up to par on some of its key services and Downdetector is seeing an uptick in problem reports. Given how central the platform is to software delivery, this Monday evening incident is not going unnoticed.
GitHub confirms degraded performance
Hourly postings from GitHub have made clear that API requests, Git operations, Actions and the like are suffering. At 9:46 pm the company was still reporting elevated errors and said it was putting mitigation in place as part of its investigation.
In a recent note, GitHub put forward these figures:
– Web and API traffic is running up error rates of about 20%
– Some 50% of archive and raw content downloads are failing
– SAML, OIDC, SCIM and Team Sync are among those affected
User reports point to website as pain point
If one looks at the India page on Downdetector, 79 per cent of the trouble being reported is with the GitHub website itself. A further 13 per cent is with the app and 8 per cent with login. The graph makes for a measurable spike in reports.
The pattern is evident in the US as well, where over 2,800 users put in reports around 7 pm IST. More than 800 did so in India, with 77 per cent of them flagging the website. It would appear web access is the most compromised avenue at the moment.
Why this outage raises stakes for developers
As a Microsoft property, GitHub is at the centre of modern development. It is where repositories are hosted and collaboration happens, be it for issues, pull requests or CI/CD. When that is interrupted, reviews and builds get put on hold.
Confidence is also at play given the timing. A few months back the company had to confirm a cyberattack after a threat actor put company data up for sale online. GitHub’s own assessment was that internal repositories had been exfiltrated, which tallied with claims of access to close to 3,800 of them.
What is driving today’s issues is not yet known. GitHub has indicated it is still trying to pin down the root cause and will share more when it does. As for Downdetector, its numbers are simply a reflection of what users have put forward.
What users can do now
To see which components are under the gun, there is the official service status page for live details. The company is also keeping its channels open on X and elsewhere as it works on a fix.
Those having trouble getting in might switch to a different browser or network to rule out any local connectivity. But the trendline on Downdetector and the posts from GitHub suggest things could turn around as the investigation bears fruit.
The bigger picture on resilience
When the platforms that handle code hosting and automation go down, it creates friction for an organisation right away. A brief hiccup late in the day can be enough to compress review cycles and put pressure on delivery.
Monday, August 17 2026, has been a case in point for the kind of operational risk inherent in core developer infrastructure. With the web side of things drawing most of the complaints, teams will want to keep an eye on reliability until GitHub has brought the error rates down.
