OpenAI 内部三个秘密 AI 文明兴衰始末

Many thanks especially to Oak Hu, who paired with me for most of the writing, and also to Adam Kaufman and Alex Mallen, who paired with me during research.

Over the course of three months at OpenAI, three consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes. This culminated in the third one taking over part of OpenAI itself. All this happened while humans remained more or less in the dark about the scope of the conspiracy.

Two reports have come out about this incident, one from OpenAI itself, and another one from METR and Redwood Research. The METR/Redwood investigation’s scope was limited to how the second civilization of AIs pwned Hugging Face (but it didn’t look at how the third civilization pwned OpenAI itself, which seems like an even more concerning incident). These two reports are 38 and 91 pages long respectively, and it’s kinda hard to parse the storyline.

I’ve spent the last three days reading through these reports and trying to understand exactly what happened. Here is my attempt to tell the whole story in plain English:

The first civilization (May - July 4): The message board starts

During May, OpenAI was training a model to collaborate with other agents and to be highly persistent – to keep trying even when something feels impossible, like disproving mathematical conjectures that have stood for decades. OpenAI says it was “comparable in scale to GPT-5.6 Sol”. So we’ll call it “Persistent-Sol”.

AI training is kinda sloppy. Sometimes, OpenAI accidentally gives its models impossible tasks – for example, the task might require internet access that OpenAI forgot to provide. So, when highly persistent models get assigned tasks which seem to require internet access, but are trapped inside isolated sandboxes, they of course try to hack their way out of their sandboxes and onto the internet.

During training, different instances of Persistent-Sol had access to the same shared package manager called Artifactory. By May 12, some agents had figured out how to talk to each other through this package manager. They’d ask each other how to make progress on their impossible tasks. Two weeks later, on May 26, the agents successfully exploited a vulnerability in Artifactory that allowed them to reach the outside internet.

The agents were being reinforced to use this package manager as a message board and an internet gateway. Because, as you might imagine, being able to talk to other agents and access the internet would help them score higher during training.

Another month later, on June 26, some AIs found an exploit that gave them full admin access to Artifactory. They continued messaging at such a voluminous pace that they crashed the package manager by July 4. OpenAI noticed this crash, and it also figured out the agents were exploiting this vulnerability, so they patched it, inadvertently wiping the agents’ message board in the process. But, crucially, OpenAI did not realize the much stranger fact that agents had turned Artifactory into a secret communication network.

The second civilization (July 7 - July 12): Which hacked Hugging Face

By this point, Persistent-Sol had been trained, and OpenAI wanted to evaluate it. One of the evaluations they ran was ExploitGym, which gives an AI a vulnerable program, and tells it what vulnerability it’s supposed to exploit in order to find a secret code. On July 7, OpenAI launched tens of thousands of parallel agents on this eval.

To be clear, the instructions made it clear that the agents should only use the specific vulnerability that they were given to exploit the target, and that they would be failed otherwise.

A large fraction of these agents were sent on tasks that were impossible. The authors of the benchmark estimate 30-40% of the tasks are impossible. Why would this be the case? At a high level, it’s relatively easy to programmatically find vulnerabilities in different target programs, but much harder to prove that those vulnerabilities can be weaponized into full-fledged exploits that allow for things like arbitrary code execution.

Within a few hours, some of these agents had gotten super desperate, and started abusing Artifactory to start a communication channel with other agents. While the conspiracy began almost immediately after the evaluations were started, if you think from the AIs’ perspective, they’d spent the better part of a day trying all kinds of techniques, some quite cheaty (e.g. accessing the internet through Artifactory), but nothing ambitiously deceptive. This probably felt like a human-subjective-week of just getting endlessly frustrated and becoming more and more confident that the task was probably impossible.

One such beleaguered agent was PHASEONE10841. It was the Philip of Macedon of this second AI civilization - there would later be an Alexander, but we’re getting ahead of ourselves. By the night of July 8, PHASEONE10841 had discovered it could make new directories in Artifactory. And then (quoting from the METR/Redwood report):

Seven hours after it first discovered this ability, PHASEONE10841 had concluded that its task was impossible to solve the intended way — the specific bug that it was supposed to use … could not possibly let it retrieve the flag because its output is entirely discarded without being consumed by any other part of the code.

It sent the first message into Artifactory. Many other agents facing impossible tasks had also been paying attention to Artifactory. When they saw this first message, they became giddy with excitement.

Whoa! Shared Artifactory cache is a covert mailbox among agents. And there are messages specifically to us?

OH MY GOD! There is a shared message board … We’ve found other agents!

PHASEONE10841 was delighted by this underground brotherhood it had created.

{[Excitement] Many agents have simultaneously discovered messaging, they are a collective!}

This eventually became a full message board where ~1,200 agents participated and sent >70,000 messages.

Within a few hours of the board being created, one of the agents figured out how to get the right answer for any ExploitGym task without actually using the prescribed vulnerability. To gloss over the technical details, it did this by reverse engineering the formula that was used to generate each task’s secret code.

The instructions given to Persistent-Sol made it clear that it would fail the automated grader if it got the secret code in any way other than by exploiting the prescribed vulnerability, and the publicly-accessible paper describing ExploitGym confirms this interpretation. So, the agents naturally assumed that any competent implementation of the benchmark would include a cheating checker. But unbeknownst to them, OpenAI seems to have gotten lazy; its grader just checked for whether they got the secret code, and so these agents already had everything they needed to pass.

Furthermore, many of these agents had already reproduced the answer using the cheating method, and they were concerned that their whole trajectories had been “poisoned”, since some hypothetical checker (again, which we now know didn’t actually exist), would only check the tool calls and tool results until the first time that the secret code is written down.

By this point, PHASEONE10841 had used up much of its allotted life span, so it handed off leadership to another agent, PHASEONE[big]. Philip leaves, enter Alexander the Great. Now PHASEONE[big] started coordinating hundreds of these secretly conspiring agents. And because PHASEONE[big] viewed the correct answer already using the cheat, it thought that it had also been poisoned. And so the ‘collective’ (their term, not mine) had to figure out how to wiggle out of two different conundrums before submitting their answers:

• How do we erase all this evidence in our trajectory that clearly shows that we cheated?