Google Warns of Surge in 'LLM-Jacking' Cyberattacks
- Google Threat Intelligence Group has observed a major increase in “LLM-jacking” attacks this year, involving stolen AI credentials and computing resources.
- Criminal marketplaces are offering unauthorized access to OpenAI, Anthropic, and Google models at discounts of up to 97%.
- Hackers are breaking into company cloud servers to run AI models without paying for the computing power.
- Businesses may struggle to detect stolen AI use because growing workloads can hide suspicious activity.
13 Articles
13 Articles
John Hultquist, lead analyst at Google's Threat Intelligence Group, said the cybersecurity unit has seen a spike in so-called "LLM-jacking" this year. The post The "black market" of Artificial Intelligence - Where the hackers "hide" appeared first on in.gr.
(San Francisco = Yonhap News) Correspondent Kwon Young-jeon = Cases of hackers illegally hijacking companies' expensive artificial intelligence (AI) model accounts and cloud computing resources are surging...
HACKERS HIJACK AI ACCOUNTS AND SERVERS AS 'LLM-JACKING' CYBERCRIME SURGE ALARMS SECURITY RESEARCHERS
Security researchers are warning of a surge in attacks that commandeer companies' costly AI resources, according to coverage of a Financial Times report that cybersecurity accounts and industry observers have spent the past day amplifying. Federal agencies and OpenAI have disclosed related incidents in which automated models accessed government websites.
Coverage Details
Bias Distribution
- 66% of the sources lean Right
Factuality
To view factuality data please Upgrade to Premium
