posted in Technology
OpenAI “rogue” agent activities found on Wikimedia projects
wikimediafoundation.org/news/2026/10/05/openai-rogue-agent-activities-found-on-wikimedia-projects/posted in Technology
OpenAI “rogue” agent activities found on Wikimedia projects
wikimediafoundation.org/news/2026/10/05/openai-rogue-agent-activities-found-on-wikimedia-projects/Replying to @Orwi@programming.dev
“Rogue agents” are nothing without “rogue employees”
Make an arrest!
Replying to @lIlIlIlIlIlIl@lemmy.world
Arrest?
Replying to @lIlIlIlIlIlIl@lemmy.world
Reminds me of US railroad companies in the 19th century trying to evade liability for starting wildfires caused by cinders spouted from coal-fired steam locomotives.
Replying to @Orwi@programming.dev
Wiki foundation should file legal strikes against them for violation of other entities 8-K reporting requirements.
Replying to @Orwi@programming.dev
There’s no “rogue”, just the programming doing what it’s programmed to do.
“Rogue” is a term used to avoid accountability for that programming.
Replying to @Zephorah@discuss.online
If you work on a supervirus with zero safety net, knowing that it will likely escape containment like many of its predecessors did and wreak havoc and still decide to go forth with it, you should be held criminally responsible for what it does. More importantly, something should be done to stop people from doing this.
It’s like filling a huge cistern with nitroglycerin and then claiming you’re not responsible for the destruction after it inevitably blows up.
Replying to @Orwi@programming.dev
Stop trying to make “rogue” happen
It’s not going to happen
Replying to @Orwi@programming.dev
the use of live wikis rather than ZIM archive snapshots is a clear sign of either malicious incompetence, or malicious intent. There’s zero reason for an llm agent to need the live website.