VULNpocalypse
HighBreach / campaignOngoing

Northflank breach leads to abuse of customers’ AI API keys, rotate all secrets

Northflank disclosed a breach on 30 Oct 2026 after Wiz flagged it; customers report unauthorised AI API key use. Treat all env vars as compromised and rotate.

Published
Updated

What happened

  • 30 October 2026: Cloud hosting platform Northflank disclosed a security incident that led to unauthorised use of customers’ AI API keys stored on the platform (BleepingComputer).
  • Cloud security firm Wiz alerted Northflank at 11:43 UTC on 30 October, and Northflank’s engineers reviewed the alert one minute later (BleepingComputer). Northflank’s published timeline records an initial network policy patch at 12:02 UTC, a request for support from the Wiz incident response team at 13:00, and further patches at 12:50 and 13:05 (Cyber Security News; BleepingComputer).
  • In an update, Northflank told customers: “Assume all environment variables have been decrypted. Disable/rotate all keys.” (BleepingComputer). Environment variables on the platform can hold AI API keys, database passwords and other application credentials (Cyber Security News).
  • Northflank confirmed the abuse, stating that “users have reported unauthorised use of AI API keys supplied to the Northflank platform” (BleepingComputer). It has not disclosed how attackers got in, how many customers were affected, which AI providers were involved, or whether other credentials were stolen or misused (BleepingComputer; Cyber Security News).
  • Northflank linked the activity to a possible wider campaign, citing IP addresses similar to those in Wiz’s “Black October for AI Developer Platforms” research, which identified at least nine incidents affecting AI development platforms in October, including Northflank, Railway, OpenCode, Warp and Lightning AI (BleepingComputer; Cyber Security News). Wiz found a shared focus on stored environment variables but has not confirmed a single common vulnerability or that the same attackers were behind all of them (BleepingComputer; Cyber Security News).
  • In two separate Wiz-investigated cases (not linked to Northflank), stolen enterprise credentials were used for about 2.5 million Amazon Bedrock API calls, generating roughly $100,000 in charges, the financial risk of “LLMjacking”, where stolen credentials run paid AI services at the victim’s expense (BleepingComputer; Cyber Security News).

What to do

Northflank’s own response and customer guidance (BleepingComputer; Cyber Security News):

  1. Rotate all credentials and secrets immediately, including those in workloads, secret groups and global secrets. If you have not already rotated user-provided secrets stored in Northflank, do so now, revoke the existing credentials with the issuing provider and generate replacements, then update the affected secrets and workloads in Northflank.
  2. Rotate AI API keys with the provider, not only those already showing abuse. Treat every environment variable as potentially decrypted.
  3. Review AI provider usage and billing logs for unexpected activity, and report suspicious usage to Northflank support and to the relevant AI provider.
  4. Disable public access to add-ons where it is enabled.
  5. Note that Northflank has enabled credential rotation for its PostgreSQL, MongoDB, MySQL, Redis, MinIO and RabbitMQ add-ons, and turned on audit logs for all customers (including free-tier accounts), review those logs.

Sources