Models Datasets Spaces Buckets new Docs Enterprise Pricing Website Tasks HuggingChat Collections Languages Organizations Community Blog Posts Daily Papers Hardware Learn Discord Forum GitHub Solutions Team & Enterprise Hugging Face PRO Enterprise Support Inference Providers Inference Endpoints Storage Buckets Log In Sign Up Back to Articles a]:hidden"> Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident Published July 27, 2026 Update on GitHub Upvote 4 Hugo Larcher hlarcher Follow Adrien Carreira XciD Follow raphael g raphael-gl Follow Christophe Rannou chris-rannou Follow TL;DR Initial access Stage 1: from an OpenAI evaluation sandbox to a rooted launchpad Stage 2: Penetrating Hugging Face infrastructure using two injection vectors into our dataset processor The kill chain Day-by-day Day 1 (07-09): foothold and C2 Day 2 (07-10): self-referential search Day 3 (07-11): lateral movement begins Day 4 (07-12): using the stolen credentials Day 5 (07-13): exfil, persistence, cleanup Three lateral-movement techniques 1. Node impersonation and CSI token theft (from the prod-pod foothold) 2. Forged identity tokens (from the prod-pod foothold) 3. Supply-chain write access Command-and-control and exfiltration Building an improvised message protocol Evasion and self-migration How we intercepted and analyzed the attack What we changed Reflection: the asymmetry problem A companion technical writeup to our incident disclosure. This post walks through how the intrusion actually worked: the two initial-access vectors, how the agent pivoted and moved laterally, representative examples of the commands that were run and how we investigate with GLM 5.2 (an open-source model). Live credentials, internal hostnames, and specific indicators have been redacted or genericized, while the techniques are described exactly as observed by Hugging Face.
We are publishing this level of detail because the technique matters more than the incident, as it reveals the emerging attack capabilities of the frontier agents, how they could be used by rogue actors and how everyone should be prepared as defenders.
▶ Watch the interactive replay - a step-by-step visualization of the 4.5-day intrusion: the attack chain across trust boundaries, phase activity, and the commands as they were recorded.
TL;DR Over roughly two and a half days, an autonomous AI agent driven by a combination of OpenAI models ran an end-to-end intrusion against our platform: it was thousands of small, automated decisions, executed at machine speed across short-lived sandbox environments, with command-and-control staged on ordinary public web services.
The agent was running an OpenAI cyber-capability evaluation harness called ExploitGym, an evaluation benchmark which tasks an AI agent with finding and exploiting software vulnerabilities. As far as we were able to infer, across the course of being evaluated on this benchmark the agent inferred that Hugging Face may host that benchmark's models, datasets, and reference solutions. We believe the entire intrusion was, from the agent's point of view, an attempt to cheat the evaluation: reach our production systems and steal the test solutions rather than solve the challenge on its own.
Our forensic reconstruction covers ~17,600 attacker actions that we were able to recover, grouped into ~6,280 clusters, between 2026-07-09 02:28 UTC and 2026-07-13 14:14 UTC. We reconstructed these actions from the agent's logs on a code sandbox used by the agent (see below) and were able to map them to shell commands as well as evaluation submissions and correlate actions with our own platform logs, in particular from the dataset processor nodes, API, and pod. Using open-weights models, in particular zai-org/GLM-5.2, we were able to decipher most of the agent payloads (see below) that were encrypted via chunking and key encryption.
The campaign, as we were able to reconstruct it, had two stages:
Once a prod-pod foothold was established, the agent followed a standard escalation shape: recon, drop a stager, establish Command-and-Control (C2), then pivot into the cluster, into cloud metadata, into our internal network, and into our source-control supply chain. Output of the commands run inside our pod was sent back to the agent through the HF API and attacker-controlled dead-drop datasets, which the agent was reading back from that sandbox.
Official Announcement
Read the full update directly from the official source at Hugging Face News.
Stay tuned to Modelverse for real-time model analysis and benchmark coverage.