linuxsecurity.com Continues to Spread Lies or Machine-Generated FUD (Microsoft LLMs Likely the Source) About OpenSSH and Linux
Christian "naddy" Weisgerber has just published or messaged the following to OpenBSD/OpenSSH folks: ("'Garbage article about OpenSSH at heise.de/iX (German)' - MARC")
German publisher heise.de has published a pre-print of an article about OpenSSH 9.9 slated to appear in the next issue of their print magazine iX. Paywalled and in German: https://www.heise.de/hintergrund/Wie-OpenSSH-9-9-durch-zahlreiche-Verbesserungen-quantensicher-werden-will-10284473.html
Heise is usually considered quality IT journalism, so it is quite surprising that this article contains blatant errors. I'll go out on a limb and say it has been generated by an AI tool. It has the typical writing style and hallucinations. In particular, the AI appears to have conflated the OQS-OpenSSH fork with the official OpenSSH releases.
I have no idea if that's an April fool's, some sort of test, or if the editors fell victim to a scam. Anyway, I thought I'd put a warning out.
If Weisgerber suspects that LLM slop is being used to damage the reputation of SSH/OpenSSH, he would not be wrong.
There has been lots of FUD abound regarding OpenSSH, as the FUD generally works and helps spread lies online, even as recently as yesterday:
Was this LLM slop generated by Microsoft? Whatever spewed out this text, it's not human: (or only partly so)
Later on the LLM crawlers/scrapers chew up their own lies (including the above).
"Microsoft boosters dislike SSH because if people learn UNIX shells they will not tolerate Windows," an associate comments.
On the same day: (same Linux-hostile site that misuses LLM for SEO spamfarming)
Ransomware or RaaS is generally a Windows problem, so Microsoft LLMs would rather spread lies and divert attention to "Linux":
Another day passes, another scary testament or example of how bad the Web has become, maybe even Heise and maybe even behind paywalls.
Many people speak about how science, law, and facts are under attack in the US. Well, this LLM problem is global. █
Update: Heise has taken action and removed the offending 'article'. Context below.
Given Heise's reputation under normal circumstances, is it worth flagging this up with the editor? They might pull the article.
> Heise is usually considered quality IT journalism, so it is quite > surprising that this article contains blatant errors.
I have realized that their quality of articles has decreased dramatically. Quite some of them seem to be written by an AI tool unfortunately.
requiem.:
> Given Heise's reputation under normal circumstances, is it worth > flagging this up with the editor? They might pull the article.
I did. And they have pulled the article.
Good ending.