I'm just reading this GNU telnetd CVE from last month. I did not realize that telnet was still a thing, but it turns out anybody could provide a username of "-f root" and, boom, they had root. The vulnerability existed for 11 years. *Wow*. https://www.cve.org/CVERecord?id=CVE-2026-24061
Don't miss this explanation of how backbone providers coordinated on this telnetd exploit in advance of the CVE release, and simply blocked port 23 traffic. https://www.labs.greynoise.io/grimoire/2026-02-10-telnet-falls-silent/
2026-01-14: The Day the telnet Died – GreyNoise Labs

On January 14, 2026, global telnet traffic observed by GreyNoise sensors fell off a cliff. A 59% sustained reduction, eighteen ASNs going completely silent, five countries vanishing from our data entirely. Six days later, CVE-2026-24061 dropped. Coincidence is one explanation.

GreyNoise Labs

@waldoj I'm not aware of any backbone provider coordination. That rarely happens for blocking anything - and probably the only time I can even recall there was such a widely coordinated port block was with Slammer over 20 years ago.

Another viewpoint here: https://www.terracenetworks.com/blog/2026-02-11-telnet-routing

Reports of Telnet’s Death Have Been Greatly Exaggerated — Terrace Networks

We see no evidence that specific core network autonomous systems have blocked Telnet, contrary to previous reports. We specifically see continued non-spoofable Telnet traffic from networks on which GreyNoise saw 100% drop-off. We suspect initial results may have been measurement artifacts or specifi

Terrace Networks
@jtk @waldoj yeah I find this rebuttal convincing too: it was a little hard for me to believe port 23 was being filtered so aggressively. The only thing I can think of close to that now is port 25 and that's mostly residential ISPs blocking traffic from their customers on the assumption it's spam.
@nelson @jtk I remember when that started, ~20-odd years ago. It was frustrating for me, as somebody running a home mail server, but I had to concede that it made sense as a spam-reduction strategy.

@waldoj

my first reaction when i read this was "who uses telnet these days" but then realized that this is probably exactly why no one was fixing bugs in telnetd.

@paul_ipv6 @waldoj

I recently heard about a major ICS/OT gear mfg that ships all end devices with telnet open and well known default creds..."for initial configuration."

@johntimaeus @waldoj

RIPE did a document with recommendations for edge devices, including not having default passwords, requiring setting a decent password before starting to route packets, etc. in the early 1990s. sad that vendors are still shipping vulnerable boxes...

@paul_ipv6 @waldoj

Grid control devices.

Of course they will never be connected to the internet because segmentation works.

@vncresolver

@johntimaeus @paul_ipv6 @waldoj That seems quaintly 20th century -right up until the law suits begin. Have they never heard of “secure by design”?
@waldoj
I suppose most orgs don’t run telnetd on their servers.

@waldoj *Wow*, telnetd has been a thing the last 11 years?

telnet client[0], sure, but telnet daemon?

[0] Yeah, yeah, there are other tools now, but my fingers have typed telnet before I've thought of something else or figured out how to spell netcat :)

@waldoj this was a bug implemented many times across many telnet daemons. the first instance was in the 80s...
@waldoj turns out, it depends on what telnet server you have. it wasn't universally true.