The most advanced part of the Chainstack breach was the weapon. The most embarrassing part was the lock it picked.
Blockchain infrastructure provider Chainstack revealed this week that it suffered an AI-powered security breach in late October that exposed customer names, email addresses and password hashes. In an updated incident report published Tuesday, the company said the intrusion is contained, customer-facing services kept running normally throughout, and customer nodes, the console and the API were not modified. It is now treating node credentials as exposed and asking customers to rotate them and change their passwords.
How it happened: five days, one default password
The timeline, as Chainstack tells it, is uncomfortably tidy. On October 26, an external attacker gained access to a development environment. On October 29, the attacker reached parts of the production environment. On October 31, Chainstack detected the intrusion and cut the access paths the same day. On November 2, it finalized the investigation and closed the incident.
The entry point was an internet-exposed Grafana instance in a development Kubernetes cluster, deployed for testing with its default administrator credentials unchanged — and without an extra layer of protection such as single sign-on or basic authentication. From there, the attacker obtained machine credentials that reached production, read secrets in bulk from the production cloud secrets manager, used a compromised VPN key to connect to the database network, and queried production databases.
The attacker accessed a database holding customer account information. What may have been exposed: organization names, account holder names, the email addresses of organization members on Chainstack, and password hashes. The attacker also had access to systems holding node credentials. Chainstack says it has no evidence those credentials were read — but it is treating them as exposed anyway, which is the right call and also the only call.
The AI part: ARTEX
Chainstack says it assesses with high confidence that the attacker used ARTEX, an open-source framework that orchestrates LLM agents to automate intrusion steps. The framework’s markers appear throughout the attacker’s artifacts, and the company links the incident to a broader AI-powered attack campaign reported in late October. ARTEX was built for authorized security testing. It is now showing up in real attacks — and Chainstack is not the only name on that list. Earlier this week, cross-chain protocol Across suffered a domain takeover before restoring control of its website.
Chainstack says it stores no payment method details such as full card numbers or security codes, so those were not exposed. It is emailing all customers and asking them to change passwords, rotate node credentials — self-service rotation is available in the console without downtime — and regenerate platform API keys. It also warned customers to watch for phishing attempts: “Chainstack will never ask for passwords, tokens, or keys.”
The Bit Take: The AI angle will get the headlines, and it should — an agent framework doing the break-in while humans slept is the new normal. But the breach did not start with a superintelligence. It started with a test server, a default admin password, and no second lock on the door. The robots are getting faster; the lesson is still the oldest one in security: change the defaults.
Sources: The Block, “Chainstack says AI-powered security breach exposed customer data; threat now contained” (2026-11-12); Chainstack, “Incident report: what happened and what customers need to do” (2026-11-03, updated 2026-11-10); Crypto Times, “Chainstack Hit by AI-Driven ARTEX Breach Exposing Customer Data and Password Hashes” (2026-11-12).
