Urgent.News

What's breaking now, across thousands of outlets.

AI

Hackers stole AI credentials to get $740,000 in free compute; Railway picked up the bill

As AI model training and distillation increase, AI tokens are becoming a piping hot currency for threat actors. Cloud hosting provider Railway found out the hard way; it has confirmed a near 26-hour battle against attackers targeting AI provider credentials to sell to AI ‘transfer stations’ for use in model distillation. The crooks generated about $740,000 in fraudulent AI usage charges using the…

Hackers stole AI credentials to get $740,000 in free compute; Railway picked up the bill

AI credentials have become a valuable commodity for cybercriminals, with one recent incident involving the theft of credentials from AI provider Railway, resulting in a fraudulent bill of $740,000. Railway's engineering team faced a 26-hour struggle against the attackers, who exploited vulnerabilities in the company's build infrastructure to access sensitive information and launch their attack.

The hackers were ultimately removed from the system on October 21, but not before obtaining the decryption keys for up to 100,000 customer workspace credentials. The attackers also managed to compromise more than 1,850 user accounts, including GitHub and OAuth tokens. The incident highlights the ease with which AI credentials can be harvested and sold, as noted by security expert David Shipley.

As AI model training and distillation become more prevalent, so does the potential for "LLMjacking," a cheaper and more scalable method for generating income compared to cryptojacking. Railway has taken steps to address the breach, including revoking and reapplying sensitive environmental variables, rotating keys, and implementing new infrastructure measures.

However, experts suggest that customers should also take proactive measures, such as revoking public network access to databases, checking audit logs, and limiting the damage that any one key can do by setting hard spending caps and usage alerts.

Written by urgent.news from Computerworld's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at computerworld.com →

More in AI

More from Thursday 12 November →