Hacker News Viewer

Together.ai: Prompt Leakage for Requests

by Fe2O3 on 11/22/2026, 3:07:27 AM

https://www.together.ai/blog/security-event-prompt-leakage-for-requests-in-june-and-july-of-2026

Comments

by: tczMUFlmoNk

&gt; During development of our public GitHub repository, xorl, <i>an AI coding agent</i> committed a credential that provided access to the private Hugging Face repository<p>My emphasis on the accountability laundering. No, &quot;xorl&quot; did not do this—<i>you</i> did. And, if &quot;xorl&quot; is <a href="https:&#x2F;&#x2F;github.com&#x2F;togethercomputer&#x2F;xorl" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;togethercomputer&#x2F;xorl</a> and `togethercomputer` is together.ai, then you wrote &quot;xorl&quot; and are doubly responsible.<p>If five years ago I had said, &quot;During development of our public GitHub repository, my computer decided to commit a credential&quot;, this would have been immediately recognized for the embarrassing abdication of duty that it is. So, too, should this be now.

11/22/2026, 4:41:19 AM


by: sans_souse

Am I inferring correctly here, in that Dragos basically had some sort of experiment set up and running for exactly this scenario, that effectively allowed them to trace the breach to the individual?

11/22/2026, 4:35:01 AM


by:

11/22/2026, 5:58:36 AM


by: nickandbro

The mass hacking continues…

11/22/2026, 4:31:05 AM