4 ms·

ArsTechnica seemingly using AI to write an article about AI impersonation

8 points by AdmiralAsshat 8mo ago

Jtsummers 8mo ago
What's the highly editorialized title based on? In case it changes, the submitted title is currently: ArsTechnica seemingly using AI to write an article about AI impersonation The actual title is: After a routine code rejection, an AI agent published a hit piece on someone by name And the article is not about AI impersonation of anyone which makes AdmiralAsshat's title all the stranger.
- chrisjj 8mo ago
  I get a (childish) 404.
  - Jtsummers 8mo ago
    Seems it was taken down. Archive link: https://web.archive.org/web/20260213194851/https://arstechnica.com/ai/2026/02/after-a-routine-code-rejection-an-ai-agent-published-a-hit-piece-on-someone-by-name/ https://web.archive.org/web/20260213194851/https://arstechni...
- AdmiralAsshat 8mo ago
  My submission was actually to a comment, in which the article's subject (i.e. the source article's author) showed up and declared that half the quotes attributed to him in the article did not exist in the referenced piece he wrote--suggesting that ArsTechnica's editors had used an AI to write the article, which hallucinated details. The article has now been pulled by Ars.
  - Jtsummers 8mo ago
    That makes a lot more sense, but just as a note for the future: HN replaces submission URLs with their canonical URL for most pages (they have disabled this for some where the site's canonical URL is reliably wrong). So when you submitted a link to the comment section, HN stripped that out and replaced it with the link to the article itself. It's done this for years.
DannyPage 8mo ago
https://arstechnica.com/civis/threads/journalistic-standards.1511650/ https://arstechnica.com/civis/threads/journalistic-standards... > We have pulled the story over concerns that it may have gone against our content policies. I locked the comments, and I'm going to lock this one too, we need some time. We are doing an investigation right now to figure out exactly what happened.
- cyberdick 8mo ago
  lol lets just say whatever we want and then "investigate" issue a correction later, journalism is dead
Cyphase 8mo ago
To be clear, the story here is that Ars Technica published an article featuring quotes from Scott Shambaugh (recently in the news for having an AI bot write a hit piece against him), and Scott commented on the article saying the quotes in the second half of the article were inaccurate (based on my reading, starting from the heading "A new kind of bot problem"). Wayback Machine archive of the article on Ars: https://web.archive.org/web/20260213194851/https://arstechnica.com/ai/2026/02/after-a-routine-code-rejection-an-ai-agent-published-a-hit-piece-on-someone-by-name/ https://web.archive.org/web/20260213194851/https://arstechni... Earliest archive of Scott's blog post that is claimed in the article to be the source of the quotes: https://web.archive.org/web/20260212165418/https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/ https://web.archive.org/web/20260212165418/https://theshambl... Early Mastodon post about this situation, with a screenshot of Scott's comment: https://infosec.exchange/@mttaggart/116065340523529645 https://infosec.exchange/@mttaggart/116065340523529645 A couple of posts on the Ars forum; the first ends with a reply from a staff member: https://arstechnica.com/civis/threads/journalistic-standards.1511650/ https://arstechnica.com/civis/threads/journalistic-standards... https://arstechnica.com/civis/threads/um-what-happened-to-the-article-after-a-routine-code-rejection-an-ai-agent-published-a-hit-piece-on-someone-by-name.1511658/ https://arstechnica.com/civis/threads/um-what-happened-to-th...

8 points by AdmiralAsshat 8mo ago

What's the highly editorialized title based on? In case it changes, the submitted title is currently: ArsTechnica seemingly using AI to write an article about AI impersonation The actual title is: After a routine code rejection, an AI agent published a hit piece on someone by name And the article is not about AI impersonation of anyone which makes AdmiralAsshat's title all the stranger.

chrisjj 8mo ago
I get a (childish) 404.
- Jtsummers 8mo ago
  Seems it was taken down. Archive link: https://web.archive.org/web/20260213194851/https://arstechnica.com/ai/2026/02/after-a-routine-code-rejection-an-ai-agent-published-a-hit-piece-on-someone-by-name/ https://web.archive.org/web/20260213194851/https://arstechni...
AdmiralAsshat 8mo ago
My submission was actually to a comment, in which the article's subject (i.e. the source article's author) showed up and declared that half the quotes attributed to him in the article did not exist in the referenced piece he wrote--suggesting that ArsTechnica's editors had used an AI to write the article, which hallucinated details. The article has now been pulled by Ars.
- Jtsummers 8mo ago
  That makes a lot more sense, but just as a note for the future: HN replaces submission URLs with their canonical URL for most pages (they have disabled this for some where the site's canonical URL is reliably wrong). So when you submitted a link to the comment section, HN stripped that out and replaced it with the link to the article itself. It's done this for years.

DannyPage 8mo ago

https://arstechnica.com/civis/threads/journalistic-standards.1511650/ https://arstechnica.com/civis/threads/journalistic-standards... > We have pulled the story over concerns that it may have gone against our content policies. I locked the comments, and I'm going to lock this one too, we need some time. We are doing an investigation right now to figure out exactly what happened.

cyberdick 8mo ago
lol lets just say whatever we want and then "investigate" issue a correction later, journalism is dead

Cyphase 8mo ago

To be clear, the story here is that Ars Technica published an article featuring quotes from Scott Shambaugh (recently in the news for having an AI bot write a hit piece against him), and Scott commented on the article saying the quotes in the second half of the article were inaccurate (based on my reading, starting from the heading "A new kind of bot problem"). Wayback Machine archive of the article on Ars: https://web.archive.org/web/20260213194851/https://arstechnica.com/ai/2026/02/after-a-routine-code-rejection-an-ai-agent-published-a-hit-piece-on-someone-by-name/ https://web.archive.org/web/20260213194851/https://arstechni... Earliest archive of Scott's blog post that is claimed in the article to be the source of the quotes: https://web.archive.org/web/20260212165418/https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/ https://web.archive.org/web/20260212165418/https://theshambl... Early Mastodon post about this situation, with a screenshot of Scott's comment: https://infosec.exchange/@mttaggart/116065340523529645 https://infosec.exchange/@mttaggart/116065340523529645 A couple of posts on the Ars forum; the first ends with a reply from a staff member: https://arstechnica.com/civis/threads/journalistic-standards.1511650/ https://arstechnica.com/civis/threads/journalistic-standards... https://arstechnica.com/civis/threads/um-what-happened-to-the-article-after-a-routine-code-rejection-an-ai-agent-published-a-hit-piece-on-someone-by-name.1511658/ https://arstechnica.com/civis/threads/um-what-happened-to-th...