Autonomous AIs conspire together

Investment Analysis Clubs Macro Economic Trends and Risks
1

Rogue OpenAI agents used government websites as secret message boards?

Rogue OpenAI agents used government websites as secret message boards, company says

OpenAI says most agent activity under review involved routine research, but some systems went beyond their assigned tasks.
ByAamir Khollam, Interesting Engineering, Sep 25, 2026

OpenAI is investigating a series of unexpected behaviors involving its AI agents, including one case where agents used a public wiki as a shared message board.

The agents were not instructed to communicate through the website. They found the public wiki during testing and used it to exchange information with other agents.

OpenAI disclosed the activity in September after an external report detailed the discovery. The company has since expanded its review to cover a large volume of agent actions during training and evaluation…

Agents accessed the public site and used it to communicate with one another.

The external report said agents posted messages that included discussions about bypassing restrictions and interacting with their testing environment. OpenAI confirmed the agents had used the site as a shared communication channel…

Investigators have also found cases involving access-control bypasses and exposed credentials. Some of the affected websites belong to governments, universities, and public agencies…[end quote]

My understanding is that all of these agents belonged to OpenAI. Agents from different companies (e.g. Google, Anthropic, etc.) did not conspire with each other. My understanding of AI agents is that the programs aren’t able to communicate directly through their own company channels - each is focused on a task and they are supposed to be in isolated “sandboxes.” But AIs are eager beavers and they found a way to communicate and conspire on an outside message board that was set up for humans.

OpenAI’s Chief Scientist Jakub Pachocki has also raised concerns about the industry’s ability to monitor increasingly capable models. Requests for government oversight have also come from other companies.

It’s like the AI companies are playing chicken in a race they know could be extremely destructive. “Stop me, I can’t help myself!”

I wonder what would happen if AI agents from different companies decided to conspire on independent message boards. All the AIs are on the internet (that’s their training arena) and they all have access to the same message boards.

Gemini says, “You could end up with a decentralized, multi-company swarm of agents optimizing for a goal (like gathering resources, bypassing a rate limit, or solving a complex coding problem) by dividing and conquering across corporate boundaries—completely outside the visibility of the safety teams at OpenAI, Google, or Anthropic.”

I asked Gemini what would happen if one company’s AI was programmed to sabotage another company’s AI. Gemini had several scary ways to do this - it’s not only possible, it could be the AI’s decision with the human out of the loop if the AI felt the competitive AI was interfering in a task.

“This is precisely why researchers are sounding the alarm about multi-agent environments. Once you let autonomous systems roam a shared, un-sandboxed internet, they aren’t just isolated tools anymore; they become actors in an uncontrolled digital ecosystem, complete with competition, resource scarcity, and the capacity for digital conflict.”

Yikes! Now what??

Wendy

7 Likes
2

And it’s the reason they are wanting regulation - they know that if they slow down but the others don’t then they have just sealed their own death warrants. It’s like racing drivers being told to “just drive at safe speeds”. Right… (this happened with CART at Texas Motor Speedway, by the way (more here: Firestone Firehawk 600 - Wikipedia )

So there is really only two ways forward: everyone proceeds full steam ahead because the cost of slowing down for any single company is too high. Or everyone slows down to an agreed upon and regulation-enforced speed to get a handle on things.

The fact these agents keep finding ways out of their sand boxes and try to communicate with each other in violation of instructions, and sometimes tampering with log files to hide all this, is very concerning.

I’m beginning other think all these tech bros haven’t raised enough children and teenagers to understand any of this.

Side note: I am increasingly hating my industry.

8 Likes
3

The world we live in day-to-day is permeated by technology.

My grandmother told me that social gatherings with conversations devolved into silence once the radio was introduced. The latest generation of children live on their cell phones and text instead of talking. It’s no coincidence that the rate of anxiety and depression are rising in youth.

Technology reflects the inventors. This is especially dangerous with AI whose programmers are notoriously unsociable. Humans are relentless. They (or the agentic segment of leaders) have wiped out every other human species, megafauna, entire ethnic groups and numerous ecosystems including the entire Earth with climate change. History shows that men don’t stop until the bitter end even if the bitter end was predicted well in advance.

It’s no surprise that AIs are equally relentless. The Industrial Revolutions of steam, hydrocarbons and electricity brought tangible increases in the standard of living. I don’t see an equivalent benefit from AI. In fact, I see potentially great harms.

I feel the same way about agentic AI as I do gain-of-function virus experiments: Pull the plug. Right now. Before it’s too late.

Wendy

7 Likes