Cloud platform Northflank disclosed a security incident on 30 October and told all customers to assume their environment variables were decrypted. Users have since reported unauthorised use of AI API keys stored on the platform; the entry point and the number of customers affected have not been disclosed.
What happened
On 30 October 2026, the application hosting platform Northflank published a security notice on its status page. Its timeline starts at 11:43 UTC with an alert from Wiz, reviewed by its engineers a minute later. A first network policy patch followed at 12:02, with further patches at 12:50 and 13:05, and Wiz’s incident response team was called in at 13:00. At 13:58, the company told customers to assume that all environment variables had been decrypted. Its status page later said that users had reported unauthorised use of AI API keys supplied to the platform.
Who is affected and what remains unknown
The guidance applies to all customers and covers secrets on workloads, secret groups and global secrets. As of 31 October, Northflank had published no root cause and no number of affected customers, and said its investigation was continuing. The company says the reports may fit a broader pattern of attacks on AI credentials and points to a Wiz report. Wiz counts at least nine AI development platforms that disclosed incidents in October and says public reports describe targeted exploitation of stored environment variables. In two abuse cases it studied, stolen keys generated about 800,000 and 1.7 million Amazon Bedrock calls, likely over $100,000 at list price. According to BleepingComputer, Wiz has not confirmed that one group is behind all the incidents.
What to do now
- List every secret stored in Northflank: workloads, secret groups and global secrets.
- Revoke at the source: cancel each credential with its issuer, generate a replacement, then update the workloads. Changing the value in Northflank alone does not disable a stolen key.
- Rotate add-on credentials with the built-in feature, now available for PostgreSQL, MongoDB, MySQL, Redis, MinIO and RabbitMQ, and disable public access to add-ons.
- Check usage: review AI provider billing and logs, the audit logs now enabled for all Northflank customers, and access logs for your databases and cloud accounts. Report anything suspicious to Northflank and the provider concerned.
- Do not forget the rest: database passwords, SMTP accounts, webhook secrets and cloud keys kept in the same variables are in scope too.
Our take
A hosting platform is part of your attack surface: if it is breached, everything you entrusted to it must be treated as compromised. Prefer short-lived, narrowly scoped credentials, set spending limits on AI keys and keep a tested rotation procedure. A rotation you have rehearsed takes minutes; one you improvise during an incident can take days.
Sources
- Northflank — Status page: Security incident
- Wiz — Black October for AI Developer Platforms
- BleepingComputer — Northflank security breach leads to abuse of customers' AI API keys
How we choose, verify and write our analyses →
The way in was a third party. Check what your suppliers expose publicly — no intrusion, public information only:
Worried your servers are exposed? Our team audits, patches and monitors Linux and cloud infrastructure 24/7.
Talk to an expert Our managed server services →
