Cloud hosting provider Railway has published a full incident report on its October breach. Attackers stole customer secrets and used them to run up about $740,000 in fraudulent AI usage. Decrypted credentials may have been exposed for as many as 100,000 workspaces.
Railway first disclosed the incident on October 22 and told customers to rotate their secrets. The new report says the attackers got Railway’s internal source code and decrypted customer environment variables for at most 1.85% of its workspaces. Railway describes that as an upper bound, not a confirmed count of stolen secrets.
How the attack unfolded
The intrusion started on October 20. The attackers exploited an unknown vulnerability in Railway’s build infrastructure to escape filesystem isolation and grab an internal credential. They used it to pull stored build histories and a copy of the source code, then studied that code for more weaknesses. A second flaw, an improperly configured trust relationship, gave them the credentials needed to decrypt customer environment variables.
Railway says the attack became a roughly 26-hour fight against attackers who used AI-powered agents to probe many routes at once while its engineers closed them. “This was not a single breach followed by a single fix,” the company said. Citing Wiz, Railway said it was defending against “abliterated models and a harness similar to ARTEX.”
What was taken
Railway says the attackers went after AI provider credentials in particular, though they also looked for cloud, database and payment keys. The stolen AI keys produced about $740,000 in unauthorized usage. Railway is working with providers to get customers reimbursed and says it will cover any charges providers won’t refund. Linked-account tokens for 1,852 users, including GitHub OAuth tokens and AI agent keys, were also exposed and have been reset.
Railway links the attack to “Black October,” Wiz’s name for a wave of attacks on AI developer platforms that also hit Northflank, OpenCode, Warp and Lightning AI. It says the attackers were removed on October 21 and found no evidence that customer Git repositories or databases were accessed.
Why it matters
AI API keys now behave like cash. Stolen keys can be resold or used for model distillation, so environment variables on developer platforms are attractive targets. Short-lived credentials, spending caps on AI accounts and fast key rotation are now basic hygiene.
Developer-platform and DevSecOps vendors can point to this incident to show the risk is real, not hypothetical. Buyers will respond better to specific claims about secret isolation, blast-radius limits and how quickly a vendor can detect and contain an attack than to broad promises of “enterprise-grade security.”
Source: BleepingComputer. Written by the xRapid Media LLP team.
About xRapid Media
xRapid Media LLP is a High Tech B2B Marketing company for technology businesses: software and SaaS companies, IT services and managed service providers, and hardware and IT product vendors. Our team combines positioning and content, account-based marketing, paid ads funnels, verified B2B data, intent signals, email and LinkedIn outreach and appointment setting to help you reach the right buyers and turn their attention into qualified leads and meetings.
Want help putting ideas like these into practice? Email [email protected] or book a time with our team.