Cyberattacks target AI’s massive compute power to mine cryptocurrency using a vulnerability in popular open source software called Ray, according to researchers at Israeli startup Oligo Security.Hackers may have breached hundreds of companies by targeting an open source software called Ray that is used to scale AI models, cybersecurity researchers have warned.
In many cases, the hackers used the exploit to install cryptocurrency miners on exposed servers, diverting the processing power used to train AI to churn out digital coins instead, according to Israeli cyber startup Oligo Security which discovered the vulnerability. In other cases, vulnerable servers were injected with so-called access “tokens” that could have allowed an attacker to breach various AI and business applications, including OpenAI and Slack.
Ray is used by some of the world’s biggest tech businesses, including Amazon, Uber and Intel, to run compute-heavy AI workloads. It’s a framework that makes it easy to spin up AI projects across distributed servers. Typically, a company will use Ray to train AI models and then deploy them in their applications. But a large number of users mistakenly left servers running Ray exposed on the internet via anthat does not require a key or password and permits outsiders to run code on the systems.
日本 最新ニュース, 日本 見出し
Similar News:他のニュース ソースから収集した、これに似たニュース記事を読むこともできます。
ソース: ABC - 🏆 471. / 51 続きを読む »
ソース: cnnbrk - 🏆 393. / 55 続きを読む »